ASIA ■~M'i ,?»^ Iff-- •^i«iSj r.fc-sS. $ :# tv;»3 Hit fyxmW Wimvmxi^ ^itatg THE GIFT OF }\ >.>A ,.. %Mf-\MXof.>y^. .JLy\. J^rn.cUo. AdWnT.. ^ ji.|./.Mi.. Cornell University Library HA 1711 1891.A32 Census of India 1891 :General report /b 3 1924 023 177 268 ,.>....« ^ The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924023177268 SALE OF GOVEftyMEJyT I^UBL ICATIQJVS. The nnder-mentioned Firms have been apfpointed sole Agents for the sale of Gtoveramant Pablications, inolnding ParUamentaiT BeportB and Papers, Acts of Parliament, Record Office PnbUoations, &o., 4bp,,and»U such works can be purchased from them either directly or through retail booksellers, who, ander the prbinsions of the sale agency contracts, are entitled to a diaooant of 25 per cent, from the selling prices:— VfSr fcll pablications eiteepiing Ordnance and Geological Maps, the Hydrographical Works of the Admiralty, wd Patent Office PnblicationB :— Me^sss. Stbb akd Sfotkiswoode, Uast Harding Street, B.C. For Ordnance and Geological Maps : — Mr. B. Stantobd, Cockspur Street, S-W. For Hydrographioftl Works of the Admiralty -.—Mr. J. D. Poitek, 31, Poultry, JLO. Patent Office Pablications are sold at the Patent Office. For all Pabliofttions exeepting the Hydrographieal Works of the Admiralty and Patent Office Publications! m SCOTLAND:— Messrs. John Mekzus & Co., 12, Hanover Street, Edinburgh, and 90, Wert Nilo Sierenb, Glasgow. IN IRELAND :— Missbs. Hodsbs, Fiooii, & Co., Likited, 104, .pralton Street, llnblin. The following is a list of some of the more important Parliamentary and other Publioations recently issued ;'— ParUauentary : Statutes— Ptiblie Ometal, Session 1892. With Index, Tables, &c., Ac. Roy. 8vo. Cloth. Price 3s. Seeond BevUed EcUUon. By authority « VoL L A,D. 1235-1718. Vol. n. A.D. 1714-1800. VoL m. A.0. 1801-1814. Vol. It. A.D. 1814- ,1880- VolV. A.D. 1830-1836. Vol. VI. A.D. 1837-1842. Roy. 8vo. Boards Price 7.. 6(i. each , SievitBd JEdHione. Tables showing subsequent Repeals, effiectfed by Acts of 55 & 56 Vict. 481^,;! Price 6d. Statutes in Force. Index to. Twelfth Edition. To the end of 55 & 56 Vict. (1892.) Roy. Bv6:" Price 10«. Staimtory Bmhs and Orders other than tho^e of a Local, Persoual, or Temporary Character. With a List of ' the more important Statutory Orders oif a Local Character ariranged in OlasB^B ; and an Index. Roy. 8to. Boards. Issued in the ]roajl«1890, 1891, and 1892. ; Price 10». each. Statutory Rules cmd Orcforafffifflloe on Ist Jannaiy 1893. Index to. Imp. 8vo. Cloth. Price 10«. Acts of Parliament, Public, Local, and Private, Session 1893, are now being publislied. Laboub Commission:-^ ' , ? Reports of the Proceedings of the three Oommittees, classed: Groups "A," "B," "0." Subjects, :^., Mining, Iron, Shipbuilding, Railways, Canals, Docks; Agriculture, Clothing, Chemical, Building, and other Trades. In course of pubUcation. [C. 6885.] iNTEENAmoirAi. llbsBjAi^t OoNiTRENCE, 1892. Report of, with Appendix. 126 pp. Price 2«. 6(J. [0. 6990.] Tkade Ukioks. Fifth Report of Labour Correspondent, with Tibbies, 1S91. Price 2».4d[. {0. 7030.] Rating Of MachisEbt Abroad. Reports by Consuls. ^ - . Price 3(^. £0. 7044.] Rahway Rates and Charges. Correspondence with regard to J^er^sed Rates; ' Price 7d. [C- 7060 and 7060.— I. apd ii., 7086 and 7098.] Eajst Indian CuBBiEjfci CommiitEb. Report, Bvidenee, &c. Complete, Price 3«. 9|d. rC. 7089.— I.] Education. IEngHANd and Wales. Report of the Committee of Council, with Annendix 1892-93, 901 pp. ,-.;,:., : Tri^Ssld. [C. 7101.] Raiiway R]?TDJiS8 OF THE United KiMJ^fiSjii 1892. 111pp. Price lid. [C. 7107.] ' Bbhrimg Sea Abbiteation. Award of the Tribunal. 16 pp. Price 2rf. [C.7137.] AGaiGUiTUEAL Education. Report on the Distribution of Grants for, 1892-93. 162 pp. Price Sid. [0.7172.] Metropoiitan Water Sotplt Commission, Report. 76 pp. Price 7Jd. H.O., No, 297/93. Buiiding Societies. Report from Select Committee on, with Evidence; Ap^ndix, &c 337 pp. Priced. 9d! H.C. Bill No. 274. District or Parish OotJNCJts, Bill to Constitute. 48 pp, Pric6 M. Obnbcs. England and Wales, and Scotland and Ieeland, 1891. Reports and Poprilation Tables are in course of publication. [C. 6986 and 6986.— i. to xii.] Mines. Reports of Her Majesty's Inspectors for the Tear 1892, with Snnunaries of the Statistical portion under the provisions of the Coal Mines Regulation Act, 1887; Metalliferous Mines Regulation Acts, 1872-1875; Slate Mines (Gunpowder) Act, 18^2, are published in thirteen separate parts. Price 6s. id. (complete). MusJBs worked in the year 1892. List of. 240 pp. Two maps. Price 2«. Abandoned Mines. List of the Plans of . Corrected to Slst December 1892. 74 pp. Price 7d' The following appear periodically, and can be subscribed for :— TkADE 01 THE UNITED KiNOPOM VlTH BRITISH POSSESSIONS AND FoBJBIOK ConSTEIBS, showins the OnftTifa'tJo. and the Declared Value of the ArticloB. K wo SJnantrtieB Trade Reports of the British Colonies, with information relative to population and general condition. Trade Bbfobts by Her Majesty's Representatives in Foreign Countries, and Reports on Commercial and General Interests. ^ The Prooeedings of Parliament in public and private business ore published daily. House of Lords Price W per 4 pp.; House of Commons, la. per 8 pp. ' State Trials, being Reports of the Chief State Trials which have taken place between 18?0 and the Dreaent time, published under the supervision of a Committee appointed by the Lord Chancellor ;— *«wom Eoy.Svo. Half cloth. Vol. L 1820-1823. (1888.) Vol. BL 1823-1831. (18iB9.) Vol. IIL 1881-1840 HRfti i Vol.IV. 1839-184S. (1892.) :. ftioM6.!S Challenger, H.U.S. A Rejxwt on the ScientiiBc Results of the voyage of, dnrioK the years 1878-1876 undar fh command of Captain George S. Hares, R.N., F.B.S., and Captain Frank Tutle^omson, ILN. PrenoMd^MidBT the Buperintandence of the late Sir C. WyviHe Thomson, Knt., F.R.S,; and now of John MufraT PhT> V.P.B.B.E. "»'M#«.y, m.i/., llis Report is now nearly complete, 39 volnmei having already been issued. EAST INDIES (CENSUS). GENERAL REPORT ON THE CENSUS OE INDIA, 1891. ^re^enteXi to Sotl^ |^ou!»e0 of ^sLvUammt tv (Urommanli of iiev Mamtv- LONDON: PRINTED FOE HER MAJESTY'S STATIONERY OFFICE, BY EYEE AND SPOTTISWOODE, ' PBINTBES TO THE QUEEN'S MOST EXCELLENT MAJESTY. And to be purchased, either directly or through any Bookseller, from EYEE AND SPOTTISWOODE, East Haeding Street, Fleet Street, E.G., and 32, Abingdon Street, Westminster, S.W.; or JOHN MENZIES & Co., 12, Hanover Street, Edinburgh, and 90, West Nile Street, Glasgow; or HODGES, FIGGIS, & Co., Limited, 104, Grafton Street, Ddblin. 1893. [0.— 7181.] Price 2s. 7d. PREFACE. Some portions of the materials for this work were so late in their arrival that advantage was taken of the delay to compress as far as possible the reyiew of the rest to within moderate dimensions, I have thus deprived myself of the opportunity, of which my predecessor, Sir William Plowden, so freely and gracefully availed himself, of showing my appreciation of the work of my colleagues, the Superin- tendents of the provincial census, by incorporating in my own review copious extracts from theirs. I am none the less under obligations to them for the help they have given me throughout the operations, and for which I now tender my best thanks. Bather than draw comparisons here, I prefer to leave their work to be judged by those who are in a position to view it from a standpoint less partial than that of one who has had so strong and intimate an interest in it and its authors as 1 have had during the last two years. I may nevertheless be allowed the personal gratification of summarising, in a few instances, what will be found expressed in a more diffuse form in the review to which these lines are a prelude. Mr. O'Donnell's work stands by itself, as in no other case has so large a population been enumerated in such detail on the responsibility of a single and practically unaided Superintendent. That his administration was statistically and financially successful is greatly due to the unsparing personal supervision he exercised on the operations from first to last. His analysis of the movement of the population, district by district, is very valuable, and his interesting iconoclastic excursions into the domain of ethnology are by no means without justification in the information he collected. Mr. Stuart was equally successful in his handling of the Madras census, and his review of the results shows a specially keen statistical scent, if the expression be allowed, in following certain paths where one has to walk circumspectly amongst dangerously fragile premises. Messrs. Gait and Maclagan have shown throughout their work the best characteristics of that most valuable speciality of India, the district officer, namely, tact, energy and knowledge of the people. Their returns were the first to be completed, and their review, especially where they respectively tread on fresh grouQd. which both had good opportunities of doing, are full of remarks worth reading. Mr. Bales, too, has written a very valuable account of the census of one of the most interesting portions of the Empire, the two divisions of Burma. Mr. Drew shares with Messrs. Gait and Maclagan the honours of punctuality, and his work on Bombay was the first to reach me in its completed form. As 1 superintended the Bombay census of 1881, I naturally feel more than ordinary interest in this part of the operations, and must frankly acknowledge that, in my opinion, Mr. Drew managed his tabulation better than I did, though in the matter of the review, I confess a preference for my own bantling. The Central Provinces were enumerated by a Superintendent who, before he went to take up the census work, had never set foot in the Province in his life. The great reduction in the cost, and the amount of information collected in the review, speak for themselves in Mr. Robertson's favour, both as administrator and reporter. Messrs. Egerton and Hastings, too, worked well in Aim^r and Berar. Among the pleasant recollections of the census, is the cordial co-operation of the officials of some of the native States, and I must specially mention the names of Mr. V. Narsinghayangar, the Superintendent of the Mysore census, whose work was one of the first to be received by me in the Simla office, and the late Mr. Bhatwadekar, and his successor, Mr. Mulshankar, in Baroda. I have also to thank, for their interest in getting the preliminary arrangements set in proper train, Mr. Tucker, / 78388. A 2 of the Central India Agency, and Mr. Erskine, in Rajpntana, as well as Mr. Grunion. in the former, and Colonel Abbot in the latter, for their work in connection with the operations subsequent to the enumeration. On the other side of the account, T must express my regret at the absence of any review of the operations and results in the second province ini the country lyin papulation. Whether the omission is due to Mr. Baillie or >hia printer I .am uiiable to say, as I have failed in extracting any explanation of, this abnormal procrastination. The reports on the census of Central India, Baroda, and Haidrabad have not been received, but the delay lias in each case been explained. •;; I do not like to conclude this long list of those directly concerned in the operations, without mentioning the. aid in connection with -the arrangement and printing of forms and tables which I received from Mr. WooUam, in Simla, and Mr. Lewis in the Bengal office, at Calcutta ; and also that given me personally by Mr. E. F. Augustin, whose services were lent me by the Military Department of the Govern- ment of India, and who, as head of my establishment, acquitted himself so efficiently that I felt entitled to specially refer to his services when replacing them at the disposal of the above Department. I must acknowledge my obligations, again, to many others, not in any way con- nected with the operations, of whose studies and experience I have taken advantage. The census deals with so many subjects, each of which, in the present day, falls within the province of a specialist, that no single individual can safely trust to his own unaided capacity in reviewing them, but is forced, like Moli^re, a prendre son bien ou il le trouve, and I have done my best to acknowledge such depredations at the time I have found it convenient to make them. The field covered, however, is so wide, that, in order to bring out the review before the information on which it is based is out of date, a mode of treatment in some degree superficial^ cannot be avoided. Oh this occasion, moreover, my time has been specially absorbed by work connected with the reprint of some of the tables, aind with the caste returns, which could not, , be completed till data had been incorporated which only reached ine in -the middle of May. It was only for four or five months, accordingly,- that I was able to. concentrate my attention on the statistics, and the amount of ground to be traversed is so vast, that I was too often reminded of the question asked of 'a late Laureate by his Quaker* friend: — "Prythee, friend Southey, with all this writing, when do'st thou find time to think?" For such views as I have been able to express, I am alone responsible. The work is, it is true, the review of a State operation, by a State official, but it is written at a distance from India, beyond reach of consultation with other ibranohes of the administration. A good deal has been added, too, in explanation of the statistics, that would have been deemed unnecessary had the work been written for official readers only, but which becomes advisable as soon as a public is approached that has no experience of what is to us in India a matter of everyday ' (observation. Whether these explanations fulfil their object, and how far we of the Census have succeeded in adequately dealing with the many important and interesting subjects thus placed within our reach, are questions that we now submit tc our readers : — Nam coenas fercula nosti'ee Malim convivis quam placuisse coquis. J. A. BAINES, India Office, London, Census Commissioner for India. 10th July, 1893. CONTENTS. Page. CHAPTER I.— Introductory and Descriptive : Physical Divisions.- -Meteorological Variety. — Areas not enumerated.— Political Divisions,. — Territorial Subdivisions {Districts, ^c.) 7-23 CHAPTER II.— Distribution of the Population : Density per Square Mile. — Comparison with Rainfall. — Density by Political Divisions . . and Physical Groups. — Congested Tracts. — Urban and Rural Population.— Distribu- tion of Urban Population. — Village Population. — House Room. — Comparison with Temperature. — Mean House Density - - - - 24-57 'CHAPTER III— Movement of the Population : General Conditions. — Special Conditions in India. — Internal Migration. — Interchange between British and Fexidatory Territory. — Rate of Increase in Population. — Comparison between Increase and Density. — Increase by Political Divisions. — Increase of Urban Population. — Increase in Material Prosperity 58-85 CHAPTER IV.— Occupation of the Population : Value of a Census of Occupation. — Scope of the Inquiry. — System of Enumeration and Tabulation, &c. — General Results. — Village Industries. — Other Occupations, classified by Orders and Sub-Orders. — Urban Occupations. — Occupations combined with Agri- culture. — Occupational Distribution by Territorial Divisions 86-120 CHAPTER v.— Ethnographic Distribution of the Population : General Sketch. — India and Burma - . . 121-130 Section A. — Mother Tongue. Scope of the Inquiry.— Influences on Language in India, Political, Territorial and Social. — Classification of Indian Languages. — Indo- Aryan Group. — Dravidian. — Kolarian. — Khasi and Gipsy. — Thibeto-Burman, &c. — Erano-Aryan. — Turanic. — Semitic and Negro-European. — Tribal Dialects compared with Tribal Population. ^Linguistic Dis- tribution of Political Divisions - - - 130-157 Section B. — Religion. Animism in India.— Brahmanism.— Buddhism.— Neo-Brahmanism or Hinduism.— Neo- Buddhism.— Sikhism.— jainism.— Brahmo and Arya Sects.— Mazdaism.— Judaism.— IsMm.— Christianity.— Minor Creeds.— Territorial Distribution of Creeds.— Sects of Christianity - - 157-182 Section C. — Caste, Tribe, or Race. General Outline.— Scope of Inquiry.— Scheme of Classification.— Principal Items in each Class and Group - - 1 ... 182-208 A 3 Page. CHAPTER VI.— Literacy of the Population : Limitations of the Inquiry. — Literacj amongst Males and Females respectively. — Com- parison of the Census with the Departmental Returns. — Literacy by Eeligion. — By Province or State.— By Age, by Caste or Race. — Knowledge of English. — Returns of University Education.— Literature ... . 209-226 CHAPTER VII.— Infirmities : Scope of the Inquiry. — Probable Errors. — Relative Prevalence. — By Age and Sex. — Territorial Distribution, — By Caste or Race ... . 227-243 CHAPTER VIII.— Sex : Territorial Variety in the Proportions. — Infanticide. — Possible Errors. — Climate and K"utrition - . . . - 244-252 CHAPTER IX.— Marriage and Widowhood : Definition of Terms. — Polygyny. — Polyandry. — Levirate. — General Features of the Indian Marriage-.system. — Prevalence of the Married. — Prohibition of Widow- Marriage. — Civil Condition by Age. — Territorial Variety in Custom - - 253-273 CHAPTER X.— Age Distribution of the Population : Errors in Return. — General Age-Distribution. — Proportion of Children. — Proportion of Adults of Working Age. — Comparison with the 1881 Return.^Computations of Life- values from Census and Mortality Returns - - 274-283 CONCLUSION.— Administration and Cost of the Census Operations 283-288 , CHAPTER I. Introductort and Descriptive. . . . . Magna modis multis miranda videtur Gentibus humanis regio, visenda que fertur, Rebus opima bonis, niult& munita Tirum vi. — Lucretius, The second general census of India was taken on the night of the 26th of February Selection of 1891, as nearly as possible 10 years after the first, which took place on the 17th of "late. February 1881. Closer correspondence in date between the two could not be obtained, by reason of the diflference in the date of the full maouj which is a factor of consider- able importance in connection with an Indian census. It is essential, in the first p]ac6> that the enumerator should have moonlight to guide him round his beat, which includes, in the rural tracts, a comparatively large area ; for his duties are not confined to the mere collection by day of a schedule which has been prepared for him by the house- holder, as in a European country. In India, the population to be enumerated contains no more than about 6 in the 100 who can read and write, so that the whole of the census record has to be prepared by special agency. As it is impossible to provide enough people of the necessary qualifications to do this in the course of a single night, the census is divided into two sections. First, the enumerator fills up the schedules of his beat with the required information for all the habitual residents therein, before the census. Then, on the appointed night, he brings this up to date by a second round, during which he strikes out all who have died or are found absent, and enters aU found on the premises who were not there when the preliminary visit was paid, such as travellers, guests, and the newly born. As he is bound to visit all the outlying parts of his beat, wherever casual sojourners are likely to be found, efficient inspection is ensured and the cost of lighting saved, by the selection of a moonlight night for the operation. But, unfortunately, the night of full moon itself cannot be fixed for the census, as it is the occasion when all the principal religious gatherings of India at shrines, temples, and bathing places are held, so that a very considerable portion of the population is absent on that night from its native place. It is a matter of uncertainty how long beforehand it is the custom to start on such excursions, which constitute one of the chief pleasures, as well as the main ceremonial observance, in the life of the masses, but, as a rule, there is a tendency to return home with greater expedition than marks the journey out. It is found, accordingly, that what with the number of festivals that are purely local in their attractions, and the facilities now available for visiting the places of wider renown by rail, the third day after the auspicious occurrence will generally find the bulk of the pilgrims back in their villages, and it was on this consideration that the census night was fixed, both in 1881 and on the last occasion. A suggestion was made, again, that the census of India should be taken synchronously with that of the United Kingdom, in order to emphasize the imperial character of the operations. The conditions of India , however, rendered it impossible to fall in with this proposal. By the beginning of April, when it was understood that ^the census would be taken in the mother country, the hot season has set in in India, ^nd the inspection of what has been above called the preliminary record, and also of the more general preparations for the census, could not be conducted, compatibly with other work, so efficiently as during the season when the supervising officials of the tract in question are on tour in the ordinary course of their duties, and it is to this inspection that the accuracy of the enumeration is mainly due. The results, showing a total of about 287,000,000 persons, were compiled Populatiop at various offices in the different Provinces and States with such expedition and enumerated, accuracy that they were published within five weeks from the census, with a difference of only five persons in every 1,000, or ^ per cent., from the finally corrected returus. The figure quoted above bears a ratio to the population of the world, as at Diversity ot present computed, of about one-fifth, and is the largest appertaining to any single circum- country with the exception of China. On the other hand, a map of the world on a ^^^^^^^ ^° A 4 Uniformity ill religion- iuid occupa- tion. Importance and variety of physical conditions. plane proiection shows that of ttie great promontories that taper from the north down into the southern seas, India is by far the smallest. It contains, in fact, but 3 per cent, of the estimated land-surface of the globe. But in spite of its political unity and comparatively small area, India, with all the diversity of its conditions, has a good claim to be considered less as a feoukliry lihan as a collection of countries, or a continent. In connection with an operation like the census, this will undoubtedly be conceded. In discussing the statistics of Europe, we distinguish, as a matter of course, those of Grreece from those of Holland, and Spain from Grermany. Even in the case of a single empire like Eussia, the circumstances of Ifkutsk are not confounded with those of Kiew or Warsaw. It is equally rational, then, to discriminate between Assam and Sindh, Madras and the Panjab, to say nothing of Burma, which has so few points of contact with the rest of this great dependency. In reviewing the results of the census, therefore, it may appear tl^t comparatively little is said of India, as a whole, for the reasons just presented, but in all cases the remarkably large numbers involved are of pre-eminent use as a centre line, from which to measure .the variations of the different component parts. But two of the facts that appear prominently in ' the returns indicate, on the surface, a significant degree of unifprmity throughout the country, so it is best to bring these to notice at once. The first, then, is that neariy three-fourths of the population is returned under one religious denomination. This matter will have to be discussed in detail in its proper sequence, but, in the meanwhile, it should be understood that the apparent uniformity is little more than a concession to conven- tional nomenclature, and the title covers creeds as well as races as fundamentally differing from each other as any in Europe. The second point to be noticed here, is the very high proportion in all parts of the country of the population living, by agri- culture. Taking it as a whole, about two-thirds, and indirectly perhaps nearly three-fourths, of the community are wholly or partially dedicated to Mother Earth, and in this case the uniformity is real, not mei-ely nominal. But it is also one which serves to throw into greater prominence the extraordinary diversity of the physical conditions under which the cultivation of the soil has to be conducted in India. We have thus to look upon the great geographic and climatic features of the country in their direct and intimate connection with the predominating means of sub- sistence, and therefore with the distribution and circumstances of the population now existing. They have to be considered, again, historically, in their capacity of determining factors in the ethnology of that population, helping to explain the variety and fusion of the elements of which it is probably composed. Some descrip- tion of the principal physical divisions of the country is accordingly necessary as a preface to the discussion of the results of their influence on its ethnology, and also on the more general subject that precedes it, namely, the ratio between population and area. The Ocean, and the Himalaya. Geographical Divisions. our- Oonsidering for the present India proper, excluding, that is, Burma, we find selves face to face with two main geographical facts, the ocean to the south, and"^e great mountain barrier to the norijh. Of the former it is enough to say that until the arrival of the European adventurers of the 16th century, it plays little part 'in the history of the population, and the settlers who came by sea are scattered and insi^ni ficant. Whether it always ebbed and flowed in its present bed, or whether India is but the remnant of the submerged continent of Leinuria, it is not necessai-v f the present purpose to determine. The question is of great ethnological momenT especially m connection with the population of the peninsular tracts of India but th evidence is of a sort that must be left to. be sifted by specialists like Messrs 'Wall o and Peschl. As regards the great mountain system of the north there ie no s '^^>h difficulty. As far back as there is occasion to look, they have stood between InT and its neighbours, more or less of a barrier to free intercourse and colonisatin^^ Planked by ranges of less height, indeed, but still difficult to traverse and backed b id and inhospitable desert, the Himalaya and it's sister systems only with great toil and hardship.to comparatively small bodies miles of bleak table-land and afford means of access '9' of men at a time. In the- extreme- north-west the way is less arduous, and has been used by all the invaders of India who were able, like Alexander and his 40,000 meni to establish, as a preliminary step, their footing in the country to the west of the ranges flanking the Indus. Through these last the passes are few and narrow, so much so, that it is said that even in the present day, a body of 100,000 men moving on India from Jelalabad, through the Khaibar Pass, would find its vanguard on the Plain of Peshawar before the rear had left the starting point. Yet it is by this route and its neighbours to the north that Iran and Turan have entered India from time imme- morial, except in the few instances when the mountain range was turned altogether, and an advance in force made up or across the Indus. At the extreme east the mean elevation is less, but on passing the range, the alternative is presented, of a com- paratively easier progress down the valleys of the main rivers of the Burma and Siam peninsula, so that much of the immigration from the north-east has probably been deflected from India to the Golden Chersonese, by the Irrawadi and Mekong. Throughout the central Himalayan system the passes are few and of great height, and it is probable that they were not attempted until after the occupation of the Gangetic basin in force by those who had found their way thither by easier routes. So far as the census is concerned, the Himalaya and its neighbours are of little importance, since, with the exception of the Kashmer State and the southern slopes, little of- the mountain tract was brought under the operations. The direct connection of the system with the present subject lies in its influence on the climatic and agricultural circumstances of the country immediately at its base, and historically, again, in its capacity of impediment to the free overflow of population from the tracts lying beyond it in the north. The former point will be noticed shortly in the course of this introduction. Eegarding the latter, all that need be said is, that though the colonisation of Upper India from abroad may have been started by bodies of immi- grants reaching the plains through the mountains of the Himalaya and Hindu Kush, the subsequent recruitment from that direction has been, so to speak, insignificant, and the expansion of the foreign communities thus settled in India has been due to causes arising within the country itself. The second of the main geographical divisions of India is the great river system, The great which covers the whole of the northern portion of the continent. The chief arteries "^^'s- are the Indus and the Ganges, with their tributaries, and the lower course of the stream we know as the Brahmaputra. The last is not, however, to be compared in importance, in connection with the present subject, to the two others, firstly, because of the competition of the great rivers of the Indo-Chinese peninsula, of which mention has already been made, and, again, because the upper portion of its south Himalayan course lies in a coniparatively narrow valley, cut off" from the rest of the country by an intervening range of considerable elevation and width. In fact, in common parlance amongst the residents of this tract, India is held to be a foreign country. As soon as the Brahmaputra emerges from its valley it becomes almost inextricably mixed up with the Ganges, forming one vast network of deltaic outlets. In Upper India the case is very different. Here the streams, which would naturally direct the course of the immigrants through the mountains, debouch at once into open country, affording the means of uninterrupted progress either south-east or south-west, for the watershed between the Panjab and the Gangetic system, once the Himalaya is left behind, is of insignificant height. The submontane tracts of the Panjab, with their continuation down the rich alluvial plain of the Jamna and Ganges, present the obvious outlet of a community disposed to agriculture, whilst the vast grazing grounds of the Indus valley are equally attractive to those of pastoral proclivities. Geographically, this belt of level country stretches all across India at its greatest breadth from west to east. Climatically, and from an ethnological point of view, it IS by no means uniform, though for perhaps two-thirds of its extent the differences are in degree only, not in kind. The third division is hardly as well defined as those already described. It The Central consists of what may be called the Second line of defence of peninsular India, that is, '"^'^S^s. the belt of table-land and hilly country that separates the Gangetic basin from the Deccan, or great plateau of the south. Its western limits are the Aravalli hills and the rainless portion of the Indus valley. On the north it rises more or less gradually from the Ganges. It is buttressed on the south by the Vindhya and Satpura hills, below Which to the south, agaiii, lies the tiongue of 'flat' country separating it from I 78388. ;3 10 The central plain and Deccan table-land. The eastern peninsular coast. the Deccan. The whole of the eastern portion is a series of broken hills from the Delta and the Orissa coastline to the central plain. Strictly speaking, the western portion, or the Central India plateau, should be distinguished from the eastern, which comprises the Chiitia-Ndgptir and Central Province hill tracts, at the Kaimlir range, but the line of separation i3 narrow and indistinct. One of the most important features of this division is its ethnological relation to the Ganges valley, on the one hand, and the plains of the south on the other, for it undoubtedly served to keep the stream of expansion of the north-western colonies within the valley, and direct it to the south-eastwards, instead of allowing its direct extension into the peninsula. So long as the fat plains along the river were available, there was no inducement to leave them in order to tempt the hardships and dangers of a trespass across the dividing ranges, so these last were left to the wild tribes, descendants of those which had been forced, off the plains by the advance of the foreigner, and here they are to be found in the present day. From the central plain, which comprises the valleys of the Narmada and Tapti, with '-what is known as the Chattisgarh tract, rises the Deccan. This title is correctly restricted to the plateau rising more or leas abruptly from the Tdpti Valley, on the north-west; from the western coast and from the southern plains, whilst it slopes upwards much more gradually from the eastern coast. The boundaries on both east and west used to be called the Grhats, or steps, but this term is now confined by geographers to the chain on the west, because the for- mation to the east consists of groups of hills more or less isolated from each other, though sufficiently in line to demarcate the table-land from the plains below and along the coast : they cannot, therefore, be called a continuous range. There is thus a very gradual transition from the conditions of the latter tract to those of the uplands, a fact in strong contradistinction to the remarkable influence exercised by the western range on the climate and rain-distribution on each side of it. There remain fo be described the strips of country along the coast, east and west. These vary in character very considerably. On the east, the upper portion is more or less an offshoot of the Ganges delta, but as the strip widens out from about the mouths of the Godavari, the country acquires a special character of its own, both climatic and ethnic. The name Carnatic formerly given to it, apparently by the early European settlers, is a misnomer, not recognised in philology or by the people. The derivation of the term has been often discussed, and, may, perhaps, be found in two words, received through the Telugu, meaning " black " and " country," or tract, a title very likely to be given by the denizens of the light soil of the eastern peninsula when introduced to the rich dark soil so characteristic of the western table-land. In the present day, the name is only given to a small strip along the western coast, and is scarcely recognised, even there, by the people, but owes its currency chiefly to official convention. The western The western coast-line is, as before mentioned, of a very different character from coasr. ^^Q gg^g^_ -pov the greater portion of its length it consists of a narrow strip of level land immediately by the shore, rising by broken ranges to the foot of the Ghdts. At the extreme south it may be taken to include the ranges of higher hills broken off at the Palghat, from the main line which supports the table-land. Towards the north, the strip widens out and becomes a tract resembling in general features the central plains, gradually giving way to the conditions peculiar to the lower Indus valley. The central part of this coast-line is almost devoid of the long-shore strip and the hills rise from the sea with but little intervening level ground, and that of comparatively low agricultural quality. The existing administrative division of Burma into Upper and Lower serves well enough for the present introduction. If the whole of the hilly country to the north had been included in the census operations, it would have been advisable to demarcate the province more minutely, and to distinguish the circumstances of the settled portions of the country, in the plains and valleys, from the hill tracts, which differ from the rest in both climate and population. Summary of The following table summarises the distribution of area and population which has geographical been attempted above, and as the administrative boundaries recognised in the census Burma. 11 I returns coincide "fairly well witli tte geographical, tlie results are generally speaking accurate. • DIVISION. A. Himalaya and Eastern Hills . . - B. Ndrthtiru Plains - . . (a) Oangetic System _ - - (b) Indus „ - - C. Central HiUs (a) Eastern Group (b) Western „ - - . D. Central Plain E. Deccan Plateau - - F. Southern Plain - - - G. N.E. Littoral H. Western Littoral J. Burma Ca) Upper (b) Lower Total, Continental India Detached Settlements (^Aden, Quettah, Andamans, GRAND TOTAL Area in Square Miles. 150,570 537,209 291,266 245,943 . 220,431 139,426 8i,oo5 97,390 193,104 62,494 30,871 96,581 171,430 83,473 87,957 1,560,080 80 1,560,160 Population. 6,542,650 151,689,676 126,627,30s 25,062,371 24,680,661 13,971,529 10,709,132 13,738,362 30,148,802 19,862,376 11,217,209 21,648,185 7,605,560 2,946,933 4,658,627 287,133,481 89,950, 287,223,431 Popu- lation per Square Mile. Per-oentage on total of 43 282 435 102 112 102 l32 141 156 318 363 224 44 35 S3 184 184 Area. 9-68 84-43 i8'67 15-76 14-12 8 -94 5-19 6-22 12-37' 4-00 ' 2-00 6-22 10-96 5-35 5-61 Popu- lation. 2-28 52-83 44-10 8-73 8-60 4-87 3-73 4-77 10-50 6-92 3-91 7-54 2-65 I-o3 1-62 It will be seen from this table that, speaking' roughly, Upper India contains 44 pel- cent, of the area and 55 per cent, of the, population with which the census is con- cerned. Southern, or peninsular India, returns 24f and 29, and the intervening belt, 20^ and 13|-. The rest, 11 per cent, of the area and 2f of the population, is found in Burma. Before entering further into the details of the distribution thus given, some Meteoro- description has to be added of their climatic conditions, for without these the logical geographical outlines serve to denote facts mainly historical in their interest, such as variety, the probable direction of colonising streams, and the demarcating lines of the ethnic divisions, which have to be discussed later. But on the climatic conditions, taken along with the general character of the tract, depends the ratio subsisting in the present day between the land and its population, since, as originally stated, the bulk of the latter lives by the cultivation of the former. The meteorology of India has occupied the attention of eminent specialists for a considerable number of years, and of late the area of observation has been widely extended by the establishment of new stations under efficient inspection, with an increase in the scope of record. There are six series of data available in connection with the present purpose, but most of them, for a period that is sufficient to afford trustworthy standards, are based on the obser- vations of only a few of the principal meteorological stations. We can make use, for instanbe, of the record of annual rainfall, of which the data are full, whilst setting aside those of the number of rainy days, the relative humidity and aqueous vapour- tension, which are available for the' principal stations diily. But as regards tem- perature, the returns alone, without the supplement of the figures of, say, the mean B 2 12 General character of the rainfall. Distribution pf the rain- fall, seasonal and geographic. Winter and spring rain. Summer rain or S.W. mon- soon. monthly range, are of little value in connection with, the subject in hand. In the course of time means will be found, no doubt, to elaborate something of the nature of a " figure of merit," meteorologically speaking, for all the most typical centres of observation in India, in which attempt every one of the above data will be given its proportional weight, but in the meanwhile the annual rainfall will suffice for the general indications which are alone relevant to the question with which this chapter is concerned. With respect to this rainfall, it must first be understood that though the mean for India of 42 inches, or thereabouts, is more or less what is to be expected in the tropics, the variations in its local distribution are so great that they are said by good authorities to be unparallelled in any Other part of the known world, -not excepting the British Islands, which boast a high rank in this respect. There is found, for example, a fall of less than five inches in the central portion of the Indus plain, and one of over 600 on a buttress of the KhAsi hills, in Assam. In the peninsula, too, along the west coast, as much as 200 inches falls 40 miles from the sea, whilst at a distance of, say, 80 miles there is only a precarious sprinkling of 20. In connection with this variability there are two points that may be noted here. First, that the rainfall is subject to much variation periodically, as well as locally; and' that from year to year the range of variation is in inverse ratio to the average -amount, that is to say, the uncertainty, so to speak, is greater as the rainfall is less. Where an inch or two makes little or no difference to the agricultural outlook of the year, the chances of a normal fall are greatest, but in the tracts wheire the said amount bears a far higher proportion to the annual total, it is most likely to be in excess or defect. The second point is that, according to the observations hitherto worked up, the same feature in the rainfall does not extend over more than two-thirds of the area under record, so that, as a rule, a deficiency in one-third of India is likely to be accompanied by a fall above the average in the rest 6f the country. It is not to be understood that failure of the rain to the extent of causing a famine is here meant, but merely the tendency, in one or other direction, of a variation from the annual mean fall. These are both important features in regard to the numerical development of the population, a subject which occupies a later chapter. But for the moment, it has to be shown how far the general distribution of the rainfall corresponds to the geographical or orographic conformation of the country. This correspondence will be found to be, in the main, very close. The rainfall in India is, it need hardly be said, periodic, throughout the country, and not liable to l)e irregularly distributed over the year. Some tracts enjoy the advantage of two seasons of rain, but for the most part the fall for the year is concentrated into one. In north-western India the season begins with the winter rains, which pass along the base of the Himdlaya from west to east, getting lighter as they advance, till they cease on the confines of Bengal. Their extension southwards, too, is confined to tlie submontane region, and, on the whole, the advantage from them is reaped in a com- paratively small tract. Megasthenes, however, judging of India from the portion of it which came most under his observation, extends the operation of this season to the whole continent as then known. It is interesting, we may remark cursorily, to find that in his time, as at present, wheat was the crop sown at this season. Before the winter rains have ceased in the north, the spring storms have gathered in the Bay of Bengal, so that the Delta of the Ganges and Brahmaputra has the benefit of them. Their influence extends, too, though lightly, down the coast, as far as the southernmost point of India, but they do not reach the western portion of the pen,insula. In their turn, they give way to the cyclonic storms so well known in the Bay of Bengal, which precede the great event of the meteorological year, the south-western rain. This sets in about the end of May or beginning of June, in two main branches. The more important sweeps up the western coast tiU it fades away towards the upper portion of the north-western plain. In its course it encounters the obstacle of the Ghat line, which impedes its progress over the Deccan, whilst in the north, the hot and dry tracts of the lower Indus induce a current from the west, which is stronff enough to overpower a weak one from the coast, and deflect it eastwards, thus depriving the upper Indus of its due rain. Moreover, it is almost certain that the 1ft coast current is aflFeoted unfavourably by a heavy snowfall during the winter, on the nucleus of the mountains we know as the Himalaya and Hindu Kush, so that political troubles are not the only ones in India that have their origin in the north-west. The second branch of the south-west current sets up the Bay of Bengal, and after irrigating Burma and the Gangetic Delta is deflected westward by the Himalayas. About the centre of the great valley it meets the other current from the west, so that when both are strong the fall is abundant, but if either, especially that from the west, is at all weak, it is only the south of the valley and the central belt, hills and plains, that receive their average amount of rain. Scarcely has the south-west current ceased in the eastern portion of India, than the Autumn rain-bearing current from the north-east sets in, across the Delta and the east coast, rain or N.E. Through this the southern plain, which receives but little rain from the south-west, "Monsoon. and that precariously, expects its main supply for the year, but the current is by no means as strong as that from the Indian Ocean, and as the latter ceases in southern India some time before the commencement of the other rain, the earth has begun to get heated again, so the extremity of the current is weakened and the fall rendered uncertain in amount. Setting aside the extreme west of India, which lies within the influence of the dry Zones of winds of the desert beyond the frontier, the minimum fall is found in the zone immediately uncertain adjacent to that region, known as the Indian Desert. But the zones of uncertainty f**!'- are to be looked for round the meeting points of the main currents, at the extremities of their respective courses. The most extensive of these lies in the centre and south of the peninsula. Here the south-western, current is stopped, as has been already pointed out, by the line of Ghats, so that comparatively little rain-bearing vapour escapes across the Deccan plateau. Again, though the current from the opposite quarter meets with scarcely any opposition of this sort, its direction lies more in line with the coast, so that only its skirts reach the high-lying land in the interior, Further south, again, this current has spent its force before it reaches the plain, and has but a small area of sea from which to replenish its moisture-bearing vapour. The fall is plentiful enough, when it comes, but it is wont to fail to a far greater extent than in most parts of India. The second of these unfortunate zones lies within the Gangetic valley, of which it includes the portion, roughly speaking, between the Indian Desert on the west and the Gandak on the east, situated mostly to the north of the main stream. In the tract under the influence of the same currents, but a little further south, the precipitation of rain, though not heavy, is remarkably regular. This is due, possibly, to the dampness of the land to the east, and to the shortness of the distance to the ocean on the west, at the point where the current catches the hills of the central belt. In connection with the relative uncertainty of the rainfall, it has been estimated from Limit of the available observations that the tracts receiving a fall averaging less than 50 inches, security of and with a mean deviation of about 12 per cent, above or below that average, all lie ^*>^^°^- ■vyithin the danger of famine. In the southern plain and the portion of the Gangetic valley that has just been described, the falls varies between 30 and 40 inches, whilst the mean range of the annual fluctuations is from 15 to 23 per cent, either above or below, or, from 30 to 46 per cent., taking both chances together. Combining the two features, geographical and hydrographic, the variety of con- Summary of ditions under which the Indian agriculturist has to labour can be appreciated. In the agrioulturnl first place, we find in the lower course of the Ganges and Brahmaputra an alluvial <'*'"<^i*i°"''- plain, abundantly watered by the fall of two rainy seasons. Higher up the valley of the former the soil is still of the same description and the country equally open and level, but the rainfall is uncertain. Keeping on towards the north-west, we enter a sub- montane plain, open and fertile, with the chance of two falls of rain in the twelvemonth, .one of which is fairly certain. Here we meet the Indus valley, and the circumstances are greatly changed. The country is still for the most part level, but the rainfall rapidly diminishes as the west is approached, so that the cultivation has to be to a great extent independent of it. The substitute in this case, as in that of Egypt, is the rise of a mighty river, which annually inundates the country for miles on each side. Beyond that range, again, and here the advantage is on the side of India, the five great tributaries of the Indus, supply the necessary irrigation wherever the level admits of^ canals. Further to the south-west, throughout the lower course of the B 3 u Indus, the rainfall is insignificant, but -wterever water can be conducted, cultivation is possible, as the soil is naturally deep and fertile. To tlie south of the desert lies the tract above called the Western Literal. The upper portion partakes of the character of the Tapti basin, in the central plain, except where it rises amongst low hills and broken ground to the central belt. But going south again, the level ground is found to be confined to the immediate coast, and inland, the country consists of series of hills trending dowawards from the Grhats. The arable land, therefore, is either on the slopes of these hills, or in the depressions between them. In the former case it is shallow in soil, in the latter, small in extent. The rainfall, however, is plentiful, though confined to a single season. In the extreme south of this strip, the arable area expands, and the rain is distributed over two seasons, so that its agricultural advantages are far greater. In the southern plain there is abundance of arable land, and the soil is, on the whole, good, though it varies much from tract to tract. But the rainfall, as has been already observed, is very uncertain. In some measure, too, the Deccan plateau resembles its neighbour. It contains a large area of light, shallow soil in the east, where the season is proverbially fickle, since this tract lies on the border-line of the two main currents affecting the peninsula. On the other hand, the Deccan has in its west and central portions considerable tracts of the disintegrated basalt now called the " black cotton soil," of great depth and fertility, with a remarkable power of retaining moisture, said to be due to the amount of alumina amongst its component elements, without which it is probable that much of the land would not be cultivable at all. Adjacent to the Deccan on the north, lies the central plain, of great fertility, but liable in its western portion to the same uncertainty of rainfall, though in far less degree, as the north-western Deccan. The hills by which it is bounded on the other sides receive, in like manner, a plentiful rain supply from both branches of the south-west current. They contain, however, but a scanty area of level ground, so that agriculture has to be carried on in more or less of a primitive scale at present, and the collection of forest produce competes with it in favour as a means of subsistence amongst the wild tribes that form the bulk of the population of this tract. The western portion of the central hill-country forms a plateau sloping in most of its extent so gradually towards the north-east that it is almost impossible to draw the ' line between it and the Grangetic basin. The strip of coast land connecting the Gangetic delta with the southern plain contains a fair extent of flat country, and as it shares with its northern neighbour the benefits of both the spring and autumn rain, in fertility it is little behind that favoured tract. The greater portion of the Himalayan tracts that come within the scope of the census operations lie beyond the region of periodic rains. The southern slopes and subsidiary ranges, receive, as rule, a -good fall, though in the north-western direction they are liable to the same weakness in the Slimmer current as the region immediately below them, in the Indus valley. Further north, the Vale of Kashmer, ^vhich is the only extensive tract of arable ground in that State, has the reputation of good and certain rain, though it is not without its experience of famine within the present generation. The country surrounding it on the west and north consists of ranges of hills and mountains, where cultivation is restricted to the valleys or the portions of the lower slopes that can be terraced for irrigation. The high plateau on the east resembles its neighbour Thibet in sterility and the absence of rain. In estimating the average fall on the Himalayan tract, the whole of this portion, Kashmer, Gilgit, and Ladakh, has been omitted, as' the observations have not been carried on long enough as yet to afi'ord trustworthy returns on which to work. Such are the circumstances of India proper, and Burma alone now remains to be described. It has been already stated that, for general purposes, the political dis- tinction of Upper and Lower will suffice, but it may as well be mentioned here that" so far as the rainfall is concerned, there is a further and a most important distinction to be drawn between the literal and the inland tracts of the lower division of the province. The hill-ranges along the coast exercise a remarkable influence in the distribution of the rain, and the fall, which varies from 170 to about 230 inches at the stations where the register is kept, diminishes to from 60 to 80 a few miles inland where our record is as yet incomplete. Again, in Upper Burma, where the fall is com- paratively light, there is a submontane and hilly tract to the north which attracts undoubtedly, a far heavier fall than the open country that forms the bulk of this part 16 of the province. Thus,' whilst the whole of Lower and the north of Upper Burma fall within the region of certain seasons, there is a considerable space in the centre which must be held to be outside the limit of immunity from scarcity of rain. Owing to the curious localisation of the rainfall in the literal tracts of Lower Burma, and to the want of accurate measurement of the areas of the subdivisions of Kashmer and Central India, these portions of India have not been included in the returns of rainfall which are discussed in connection with the distribution of the population in the next chapter. Omission of certain tracts from the rainfall statement. The last subject that calls for notice in this introduction is the division of the Omissions country politically and for administrative purposes, but before entering into this, it is ^'■°™ ^^^ as well to explain exactly how far the census operations extended, and what tracts ^^„"^ ^^' that fall within the general delimitation of India have been omitted from the enumeration. In the first place, the chief Himalayan State of Nip^l is excluded, and so has the corresponding State of Bhdtan, further east. The small State of Sikkim, which lies between these, was only enumerated informally, though as nearly as possible synchronously with the census of the rest of the country. The population is not, however, included in the general returns, as no details beyond the actual number of each race were recorded. The same process was adopted with regard to the Cis-Salwin Shan States, on the east of Burma, and in the rural parts of British Balochistan excluding, that is, Quettah, Lorelai and the railways. In the case of Manipur and of a few tracts along the northern Burma frontier, the census was taken, but in subsequent disturbances the records were destroyed. In 'the Andamans,' beyond the limits of the convict settlement of Port Blair, in the Kakhyin country, bordering on Burma and Assam, in the Trans-Salwin Shan States, and in a wild tract in the south of Rajputana, no enumeration was attempted, though in the last case a registration by households was recorded. The marginal note shows the approximate population of these tracts. There are also to be considered the small posses- sions of Portugal on the west coast and the detached settlements of Prance, which are enclaves in the provinces of Bengal and Madras. By the courtesy of the French officials a census of the latter was taken simultaneously with that of British India. The population of the Portuguese territory here shown is that of the census of August 1887, which was published in October 1892. The normal population dealt with then in the succeeding review is 287,223,431, but the full detail required in the standard schedule was not prescribed for the less advanced tracts, so that from the normal figure shown above there will be found deductions made in all the more elaborate returns, and due explanation of the difference has been appended to the tables concerned. The results of the census are shown by political and administrative divisions, on which some explanation is necessary before discussing the statistics in detail. The first subdivision is into British and Feudatory territory. The former is the dominion of the Grown, under the sole administration of the British Government, and contains 62 per cent, of the area, and 77 per cent, of the population now dealt with in con- nection with the census. The remainder is under the rule of the Native" Chiefs, subject to the advice and control of the British Government, in its capacity of paramount power, and includes 38 per cent, of the area and 23 of the population. The higher proportion of the area in this category is due to the inclusion in some of the States of wild tracts like the Himalayan regions of Kashmer, the desert portion of the Indus valley, and nearly all the forest-covered hills of the central belt. The first-named lies outside the turmoil of the dynastic struggles that troubled Upper India for so many years, and after its brief experience of Sikh rule, was made over by the British to the chief of a warlike clan of the outer Himalaya. The general sterility of B 4 Sikkim {Registered) Manipur {Estimated) ... British Balochistan {Registered) Cis-Salwin Shan States {Registered) Burma Frontier tracts . . - Eajputana Hill tracts {Registered) 30,458 250,000 145,417 372,969 116,493 204,241 Total excluded 1,119,578 French Settlements . - . Portuguese Possessions ... 282,923 561,384 Total, Foreign 844,307 Included in the Census 287,223,431 Grand Total 289,187,316 Political divisions. British and Feudatory. 16 the territory, with the exception of the Vale of Kashmer, which, as observed above, is of but restricted area, and the southern slopes of the hills that separate it from India, has been already mentioned. Rajputdna, again, a country of large towns in an open plain, was obviously selected as its stronghold by a warlike race, which let the tide of invasion from the north-west roll along the frontier down the rich plains of the Ganges unheeded, so long as their desert home was unmolested. The plateau that forms the western portion of the central belt was carved out into estates by various adventurers from diflFerent parts of India, in the disturbed times preceding the establishment of the British dominion, and only the eastern and wilder portion contains dynasties of ancient lineage, strong in their mountain fastnesses. The eastern section of this tract of hills is split up into many petty chieftainships, as might be expected from the broken and difl&cult nature of the country. It offers no attraction to the civilised denizen of the plain, so it has been left to the scattered remains of the former occupants of the latter, driven off it to take refuge in a land that like Cameliard, before the coming of Arthur, is " Thick with wet woods, and many a beast therein. And none or few to scare or chase the beast." On the other hand, the alluvial plains of the Ganges valley, and the equally fertile strips along the northern coast on each side of the peninsula, fall, the one entirely, ^he other almost so, within British territory, and contain in the aggregate more than half the popnlation of India. In size and population there is no more uniformity amongst the Provinces and States of India, than amongst the political units of Europe. The British possessions were built up gradually, and with regard, as far as possible, to the demarcation in force under the government that preceded the. acquisition. Within the last generation, at all events, the transfers of territory have been comparatively trifling in extent. Amongst the more important are those of the Delhi tract to the Panjab from the Agra government, or what is still called by the clumsy and obsolete title of " the North- West Provinces " ; and of the Surma valley from Bengal to the then newly-formed province of Assam. In the south, again, a small tract has been detached, for adminis- trative convenience, from the Central Provinces, and added to the Presidencv of Madras, whilst the still smaller Jhansi territory was exchanged by Sindhia in 1886 for the fortress of GwaHor.: The Native States exist under the guarantee of the paramouut power, so their territorial limits are as they were found at the time when they entered into their agreements with the British Government. As the principle of guarantee is the same for small and great, these States vary in size from the dominion over a village or two, with a population of a few hundreds, to the lordship of *more than 80,000 square miles, with nearly 12,000,000 of inhabitants. For convenience in respect to their relations with the paramount power, the smaller States are grouped into Agencies, a term applied to even those of larger size, which are either much intermingled territorially, or which are of so homogeneous a constitution that a single intermediary between them and the Supreme Government is suflBcient, In both of these cases there are subdivisions of the authority of supervision, delegated to local Agencies or Residencies, but though these are recognised in the special census returns for the Agency as a whole, they do not appear in those under consideration, which deal with the aggregate as the unit. Central India, Balochistdn, and Rajputdna are the main instances of this. The smaller groups of States are combined in the Imperial returns according to the local government or administration with which they are politically connected. For example, those attached to Bombay appear in a single item in the general returns, and under 19 heads in the provincial volume, but individually they number several hundreds. It is as well to note here the distinction between the position of States such as these and that of tracts ,in Bengal, Oudh and parts of Madras, on the owners of which titular honours, similar to those of chiefs may have been bestowed by the British Government. In the former case there has been the recognition of distinct political entity, whereas the second case is merely that of an estate forming part of British territory, and with nothing but a special fiscal agreement to distinguish it from the rest. . The following table shows the area and population of each of the British Provinces and States, or, as explained . above, groups of States. It also, includes, a' statement of 17 the proportion borne by each to the total area and population respectively of India as a whole : — Percentage on Total. Province, State, or Agency. Area in Square Miles. Population, 1891. Area. Population. Bengal 151,543 71,346,987 9-71 24-84 Madras 141,189 35,630,440 9 05 12-40 f N.-W. Provinces \ Oudh - 83,286 34,254,254 5 a4 11-93 24,217 12,650,831 1 55 4-40 Panjab 110,667 20,866,847 7 09 7-26 f Bombay - \ Sindh - - 77,275 15,985,270 4 95 5-56 47,789 2,871,774 3 06 1-00 Central Provinces ' - 86,501 10,784,294 5 55 3-75 ' Upper Burma Lower Bur nut 83,473 2,946,933 5 35 1-03 87,957 4,658,627 5 64 1-63 Assam 49,004 5,476,833 3 14 1-91 Berar 17,718 2,897,491 • 1 11 1-01 Ajmer 2,711 542,358 17 0-19 Goorg 1,583 173,055 10 0-06 (Aden 80 44,079 0-01 0-02 1 Quettah, &c. — 27,270 — 0-01 [_ Andamans — ■ 15,609 — — Total British Provinces 964,993 221,172,963 61-85 77-00 Haidrabad 82,698 11,537,040 5-30 4-02 Eajputana 130,268 12,016,102 8 35 418 Centraf India 77,808 10,318,812' 4 99 3-59 Mysore 27,936 4,943,604 1 79 1-72 Baroda 8,226 2,415,396 53 0-84 Kasbmer 80,900 2,543,952 5 19 0-89 States connected with Bombay 69,045 8,059,298 4 42 2-81 ,, „ Madras 9,609 3,700,622 62 1-29 „ „ Central Provinces - 29,435 2,160,511 1 89 0-75 „ „ Bengal 35,834 3,296,379 2 30 1-15 „ „ N.-W. Provinces - 5,109 792,491 32 0-28 „ Panjab 38,299 4,263.280 2-45 1-48 Fort Steadinan, Shan Outposts — 2,992 — — Total, Feudatory States 595,167 66,050,479 38-15 23 GO GRAND TOTAL, INDIA - 1,560,160 287,223,431 100-00 10000 Their distribution over the physical divisions described above is shown in the table given at the end of this chapter. The figures can also be compared with the corre- sponding returns for countries more widely known, so that an idea more vivid than a mere statement aifords may be obtained of the relative size and population , of the different parts of India. Bengal, then, to begin with the largest of the provinces, has an area equal to that of the United Kingdom, with the addition, of a second Scotland, whilst the population is about that of the whole of the United States of America in 1890, too-ether with that of Mexico. The Presidency of Madras has the area of Prussia and Saxony, and contains a population exceeding that of those two States by that of Wurtemburg, one of the smaller members of the German Empii'e. If the native States under this government be added, the aggregate population slightly exceeds that of the United Kingdom or of Austria-Hungary, at the last census. The North-West Provinces, with Oudh and the connected States, supports nearly the same number of people as are found in the whole German Empire, though the area is something under that of Italy. Sindh and the Panjab are about equal in population to Austria, and with the States connected with the latter, to little below England and "Wales. Similarly, Bombav and its States are comparable to Spain, Holland ' and Norway. The whole population of Brazil, with its 3,200,000 square miles of area, could be accommodated, with room to spare, in the Central Provinces, which extend over no more than 86,500 miles. Assam, on the otiier hand, shows as many people as Bavaria, but on nearly twice the area. The island of Oeylou returns about the same population as Sindh and. Berar, respectively, but the former contains nearly twice the area, and the latter little more than two-thirds. The population of Lower Burma and of Ireland could exchange places, but the former would be woefully cramped in the Emerald Isle after the enjoyment of more than twice the area in its native peninsula. In the acquisition of the rest of Burma, the second biiJe of the cherry, which Lord Dalhousie prophesied we should t 78388, 18 The smaller Statest Bombay.. Madras States. Central Province and Bengal Stales. North- West Province States. Punjab States. have to take, tiiere has been added to the Empire the population of Switzerland, on the combined area of Portugal, Greece, Holland, and Belgium ; or nearly the area of the colony of Victoria, with more than three times the population. ^ In the aggregate, the British territories cover an area equal to that of the whole of the Teutonic countries of Europe, with the addition of Hungary, Servia, and Bulgaria. There is also the Feudatory territory, with more than the population of the United States, and with an area equal to that covered by the Triple Alliance, with Belgium and Servia thrown in. The two great agencies of Rajputana and Central India alone, extend over the area of the German Empire, and contain nearly the population of Austria. Haidrabad is as large as England and Scotland put together. Mysore is a little smaller but more populous than, Portugal; and, to compare mountainous countries together, Kashmer shows the population of Chile, but on less than a third of the area. Next to these large States or Agencies, come the groups connected with the various local governments, some of which deserve a few words of description before the general subject is dropped. The Bombay States form the largest group. Amongst them, the small State of Khairpur, still in possession of the late ruling family of Sindh, and the peninsula of Kachh, appertain to the Indus valley rather than to the rest of the Presidency. Then comes the peninsula of Saurasthra, now known as Kathiawar, from one of the dominant tribes froni the north that immigrated thither in years, gone by. Under this general title are included many small States and four of con- siderable size, Junagadh, Nawanagar, Bhaunagar, and Drangadhra. On the adjacent mainland, bordering on the western hills of the central belt, are four or five smaller groups of much the same constitution. The rest of the coast strip falling within this Presidency contains a coupl^ of small States, interesting from their connection with the former ruling races of the interior. One was the appanage of the Sidi, or Abyssinian commander of the Moghal's fleet. The other was held by a maritime chief of the Maratha race, not free from the imputation of piracy. Not to mention a few wild States imbedded in the forests at the foot of the Ghats, where these mountains fade into the central plain, the remaining States in this part of India are mostly creations of the Peshwa's rule, with the addition of Kolhapur, descending from Shivaji, the founder of the Maratha dynasty. To the south of Bombay come the States connected with the Madras Presidency, of which there are but five. Three of them are merely small, enclaves in British territory. The remaining two lie along the most fertile portion of the western coast, and the larger of them, Travancore, is the more interesting, owing to the fact that from its isolated position it has been enabled to conserve features of racial and social development long since obliterated in the busier life of the tracts repeatedly overrun by foreign influences. The small States connected with the Central Provinces, and also those attached to Bengal, have been guaranteed, for the most part, to the descendants of some of the chief forest tribes, displaced from the plains by various waves of more civilised immigrants. Nearly all of them are situated in the hilly country composing what has been called above the central belt, in its eastern section. In Bengal, however, there are two detached States, that of Hill Tipperah, under one of the tribes of the' eastern frontier, and that now called Koch-Behar, the appanage of a more northern tribe, which, since its settlement in the lower valley, has become incorporated in the general population of the Gangetic delta. There is a wide expanse of British territory between the above States and the next, which are connected with the Government of the North-West Provinces. These last are only two in number, and detached from each other. The smaller but more populous Rampur, lies imbedded in British territory. The other, Garhwdl covers a large area, sub-Himalayan in character, between the Panjdb and the North- West Provinces. After Bombay, the Panjdb Government has the largest and most populous collection of States connected with it. They can be divided into three sections. First, those of the plains, the largest of which, Patiala, Jind, and Nabha, with a few others, are known as the Sikh States. Then follows the large but sparsely peopled tract under Baha- walpiir, on the Indus, between Sindh and the northern portion of Raiputana Last come the Hill States, stretching some way into the Himalaya. These are 23 ' number, but most of them are comparatively insignificant, and though in th aggregate, of considerable area, contain but a small population. The largest' Ndhan or Sirmiir, is the best known, perhaps, from, having given its nam,e to one of the stauuuhest of the regiments of Hill-men that fought in the Mutiny. '' ' 19 In the table accompanying this cliapter, which shows the distribution of tlie Table on political divisions over the divisions based on geographical and, climatic considerations, ^^^^ ^^" the territory under feudatory rule has, been combined with British territory, wherever it is not recognised in the general census returns as a separate unit. ■ British territory. The only topic that remains to receive notice in this introduction is the adminis- The "dis trative subdivision of a Province. In the case of States, enough has been said to r"?*.'!,'" indicate the very varying character of the unit as to size and constitution, whilst the larger States are not subdivided on any uniform system. In British territory the diversity is mainly in size, as the authority and duties of the officer placed in charge are everywhere the same, in so far as responsibility is in question. From an administrative point of view , the head of a State is not assumed to be bevond the influence of bad advisers, whatever be the range of his sovereignty. In Russia, again, it is said that when anything goes wrong, investigation begins at the bottom, and the tendency is to throw the responsibility upstairs, as far as audacity can carry it. But in British India, experience has dictated the judicious middle course of combining considerable power with full, responsibility over an area and population within the compass of one man's efficient supervision. In detail, the size of the district differs from province to province mainly according to the system of administering the land and its revenue ; but, even within the province itself there is much variety, as can be seen from the following table : — — I — ■ Extremes. Number of Mean of Districts. Province. Area. Population . Districts. Area. Population. District. Area. District. Population. Madras* 24 35-4 5,882 42-4 " 1,465,747 • Nilgiri Vizagapatam Agency 957 12,623 Nilgiri Malabar 99,797 2,652,565 38-0 27-5 r Broach 1,463 Panoh Mahals 313,417 Bdmbayf 18 4,292 842,417 • Khandesb 10,907 Khandesh 1,460,851 3z-6 47" 7' ■ Upper Sindh Fron- 3,549 Up. Sindh Front. 174,548 Sindh - ' - 5 9,558 574,355 ■ tier. Karachi 14,182 Haidrabad 918,646 43-1 38-1 " 1,503,520 ■ Howrah 476 Chittagpng Hills 107,286 Bengali 47 3,224 Lohardaga 7,140 Maimansingh 3,472,186 34-5 31-9 1" ' Tarai 963 Dehra Diiii w 168il35 N.-W. Prov. and Oudh 49 2,194 957,247 ■ Kumaon 7,151 Gorakhpiir 2,994,057 46-3 3o-i ■ Simla 102 Simla - 44,642 Panjab 31 3,570 673,124 • KAngra 9,574 Si51k6t 1,119,847 38-8 42-6 Narsinghpur 1,916 Nimad - 253,486 Central Provinces 18 4,806 599,127 ■ Raipur 11,724 Raipur 1,584,427 55-3 52-5 r Kyank-s6 1,050 Ruby Mines 26,134 Upper Burma 17 4,910 173,349 ■ Upper Khyindwin 19,000 Mandalay 374,060 47-0 52-4 r North Aralian 1,015 North Arakan 14,628 Lower Burma§ 18 4,885 248,795 i Amherst 15,203 Bassein 475,002 3i4*6 73-5 r North Kachar 1,728 North Kachar 18,941 Assam 13 3,769 421,295 i Khasi Hills 6,041 Sylh^t 2,154,593 10 -8 i8-3 r Elichpfir 2,623 Elichpur 315,798 Berar 6 2,953 482,840 < Wdn 3,911 Amraoti 655,645 r Simla 102 North Arakan 14,628 Total, Provinces -, 24911 3,875 880,965- Upper Khyindwin 19,000 Maimansingh 3,472 186 * Including Madras city, and taking the three Agency tracts as distinct units. t Excluding Calcutta and two suburbs. 11 Including the three districts of Coorg and Ajm&e-Mervara. C 2 t Excluding Bombay city and Aden. § Excluding Rangoon city. 20 The exponential figures above each item represent the percentage of the mean variation from the arithmetical mean of the province as a whole. For example, tHe average area of a district in Bengal is 3,224 square miles, but owing to the number of districts of considerably greater or less area, there will be found, on the whole, a difference of 43 per cent, between that figure and the area of any single district. On the other hand, in Berar, where the number of districts is small, and they are more uniform in size, the mean difference of the average is less than 11 per cen.. Such additions to an average are of use in so far as they indicate whether the mean figure is typical within the province, or, on the contrary, is no more than an arithmetical expression, serving only to compare one province with another. It is clear that the figures in the margin belong to the latter class, for even in Berar we find the area of the smallest district is 2,623 square miles, and that of the largest 3,911. But the more valuable comparison is between the average of one province and that of another, and it is also interesting to set beside these averages the areas and populations of better-known tracts, such as those of an English county. For instance, the average Indian district covers an area just below that of Devon and Cornwall, or the four Home counties, whilst its population is about equal to that of the three northern counties of Northumberland, Cumberland, and Westmoreland, with that of little Huntingdonshire thrown in. Of the provincial averages, and taking into consideration both area and population, it will be seen that the heaviest charge is to be looked for in Madras, where, if the three Agency tracts are considered to be separate units, the district is about the size of Yorkshire, or of the Grand Duchy of Baden, with the population of Nottingham and Durham. If we take the Agency tracts to be part of the district to which they are subordinate, which can fairly be ^done, we raise the hvea to nearly that of Wales, minus Anglesea, with a considerably larger population than that of the " gallant little " kingdom. The provincial average that comes nearest to that of the country at large is found in Bombay, which approaches in area the three south-western counties of Somerset, Devon, and Cornwall, with a population that lies between that of Devon and Cornwall, on the one hand, and Somerset and Cornwall on the other. In some cases it appears that the discrepancy between area and population is very large. For example, in Upper Burma, the population of Tyrone is found scattered over four times the area. In the lower section of that province the area is much the same,, but the population rises to more than that of Tyrone and Fermanagh together. The same disproportion is observable in Sindh, where, on the area of the four northern English counties, plus the West Riding, there is a population equal only to that of Westmoreland and Northumberland, or within a little of that of New Zealand. The contrary state of things is seen in Bengal and Oudh, where the district is small in size, but contains a very large population. The average Bengal district compares very well in both respects with Norfolk combined with Staffordshire, and one district, we may remark, in passing, contains almost as many people as the Kingdom of Saxony, within far less space. The North- West Provinces, taken with Oudh, as in the table under consideration, show an average of the area of Norfolk, and twice the population ; but Otidh alone shows nearly the population of Staffordshire on the area of Northumberland. Half Wales in area, with the population of Glamorgan, about represents the Panjab district, and that in Berar would just overlap Gloucester and Somerset, but its population is scarcely that of the latter alone. The former furnishes in population a close parallel to a Central Province district, which, however, has nearly four times the area, as is the case with Burma. With this the list closes, and enough has been said to give some notion of the dimensions of the district unit, though there is nothing in the description that serves to delineate the great importance of the position held by this division of territory in the adminis- trative scheme of the British dominions in India. A few words, however, will not be out of place as to the part played by the district in the census operations. In the first place, it is the unit of compilation in the provincial returns, so the tables of specific population in the next chapter and of the movement of the population in that which follows it are all based on the district figure. Still more to the point is the fact that all the details of the arrangements for taking the census and for providing agency for the enumeration and also, in most provinces, for the tabulation of the results were left in the hands of the district ojBBcer, acting under the instructions of the Provincial Superintendent of the census operations, so that the accuracy of the work was mainly dependent on the efficiency of his supervision. It was directly from his hands that the preliminary totals were received in the central compilation office for India, at Simla, and as it has been already mentioned that these last were furnished in time for publication five weeks from the date of the census, it is super- fluous to add comment on the admirable co-operation and energy that characterised the connection of these officers with so heavy an addition to their current duties. 21 GEOGRAPHICAL DISTRIBUTION OF THE POLITICAL DIVISIONS OF INDIA.* * See page 17. C 3 22 Geographical Distribution of the [GomplementWry to the Table Total of Province and States. PROVINCE AND STATE. 1 Area. Population. 1, 2. 3. Panjab and States 148,966 25,130,127. Kashmer 80,900 2,543,952 N. - W. Provinces, States, and Oudh. 112,612 47,697,576 A. Himalaya and Eastern Frontier Hills. B. Northern Plains. B. (1) Gangetic System. Area. Bengal and States Assam Central Provinces and States. Central India Ajmer-Merwara Rajputana Bombay and States Baroda Sindh, IKhairpur and Kaehh. Berar Hyderabad Mysore Coorg Madras and States Burma Detached Settlements 187,377 49,004 115,936 77,808 2,711 130,268 133,711 8,226 60,398 17,718 82,698 27,936 1,583 150,798 171,430 80 Total India - 11,560,160287,223,431 74,643,366 5,476,833 12,944,805 10,318,812 542,358 12,016,102 23,354,216 2,415,396 3,562,126 2,897,491 11,537,040 4,943,604 173,055 39,331,062 7,605,560 89,950 Population. 20,615 80,900 1,645,364 2,543,952 18,137 1,380,376 10,669 20,249 Area. Population. B. (2) Indus System. C. Central (1) Eastern Group. Area. 21,346 5,642,878 468,042 504,916 94,475 109,448 28,755 2,711 34,531 46,317,200 62,919,111 4,971,917 107,005 Population. 17,841,885 542,358 6,233,841 I Area. { Population. 10. 11. 59,088 47,171 14,000 150,570 6,542,650 291,266 78,540 60,398 3,658,360 3,562,126 7,378,819 3,564,785 1,733,581 126,627,305 245,943 25,062,371 19,167 139,426 1,294,244 13,971,529 23 Political Divisions of India. given on page 11.) Hills. (2) Western Group. Area. 12. Population. 81,005 13. 63,808 8,585,231 17,197 2,123,901 D. Central Plain. Area. 14. Population. 15. 68,765 9,380,020 10,907 10,709,132 17,718 97,390 E. Decean Plateau. Area. 16. Population. 17. 1,460,851 2,897,491 13,738,362 52,028 F. Southern Plain. Ai'ea. 18. Population. 9,649,864 82,698 27,936 1,583 28,859 193,104 11,537,040 4,943,604 173,055 3,845,239 62,494 30,148,803 ' 62,494 C 4 19. G. North-East Littoral. Area. 20. 8,172 Population. 21. 3,877,294 H. West Littoral. Area. 22. Population. 23. 19,862,376 19,862,376 22,699 30,871 70,776 8,226 12,243,501 2,415,396 7,339,915 17,579 6,989,288 11,217,209 I 96,581 21,648,185 24 CHAPTER 11. Distribution of the Population. Xaoi 8' epya. x€§/crTe7tXou(rii/ eKrjXoi. Ile^Oi' e'v aAXoT§tai(r< ^oai/ i (TTurraTo kcufi/xig, — Theokritus. The population of India, which was roughly set forth in the statements incor- porated in the last chapter, has now to be considered in its relation to the land, and again, with reference to its social distribution, into rural a,nd urban communities. As regards the former, the average number of persons to some conventional unit of area, such as the square mile, affords the best means of comparing one country or tract with another, with respect to the general distribution of its inhabitants. But we must discriminate between the density, as this ratio is called, and the pressure of the population. A low specific population may imply pressure, as much as a high one, and, conversely, great density is not incompatible with a high standard of comfort. We need not look beyond India to find two tracts with an equally small number of inhabitants per mile, m one of which the sparsity is due to some physical cause, and may thus be considered more or less permanent ; whereas in the other political uncertainty alone has impeded the exploitation of what is by nature capable of supporting eight or nine times the existing population. Again, we see people crowded to a degree that in a rural community is almost incredible, but proving by their continued multiplication that the margin of subsistence has not yet been reached, whilst in other tracts, about the same specific population is only maintained by the habitual importation of food supplies from outside. These last are, however, very rare in India, where the bulk of the population is engaged in the production of food, and the preponderance of mining and manufacturing industries that is found in England, Saxony, and parts of Belgium, is unknown. The most densely peopled tracts of rural India are, as a rule, no worse off than the denizen of happy Auburn, for whom " Light labour spread her wholesome store ; Just gave what life required, but gave no raore.'' But there is no doubt that the region of periodic rainfall is not the place for the experiment of hov\- fine the provision of nature can be run, and apart from the tracts mentioned above, which are habitually not self-supporting, there are others in which the return from land is so near the economic limit that little more than the year's supply of food is to be expected from it. If, then, they fall within the zones of uncertain rain, described in the introductory chapter, a large number of people have to resort to the market as purchasers, and though the greatly iucreased facilities for transport have tended to equalise prices, there is no doubt that the normal rates have been raised throughout the country, and the agriculturist has to face the change from another point of view than that which he occupies in a good year. But even in the parts of India thus unfavourably situated the amount of capital put by is remarkable. As in the days of Bernier, it is chiefly in the form of gold 'and silver ornaments; but these are now more apparent than when he wrote, irrespective of the relative amount, for it was then the object of every man to conceal the fact that he was other than a pauper, whereas uow-a-days, the pemlulum has swung to the other extreme, and the police courts show too frequently the temptations put before the criminal classes by the display of ornaments of value on the person of women and children when on journeys and in other comparatively unprotected situations This form of wealth is generally kept in the family, and it takes quite exceptional circum- stances to bring it into the market. A striking instance of this was afforded durino- the hard times of 1876-78 in southern and western India. In the early yeai's of tha't decade, barring a short period in 1871, the average monthly value of the ornaments sent to the Bombay Mint was only Rs. 6,000. A rise took place at the end of 1876 and continued till it reached Rs. 1,900,000 in September 1877. With the return f good times to the south the amount decreased, but in 1878 the scarcity in some of tl° Fative States in Gujarath raised it again to about Rs. 7(''0,000 till the end of the ' The rarity of these manifestations of distress indicates that the landed class generaHy 25 raises crop enough to cover what may be termed normal fluctuations. There is also to be borne in mind the fact that the system of village finance adapts itself very fairly to the circumstances with which it has to deal, and often helps the raiat to tide over a crisis. In the first statement given in the introduction there is a column showing the General mean number of persons per square mile in each of the chief physical divisions. It density. was inserted in anticipation of the present explanations, in order to give some general notion of the difi'erent circumstances, in this respect, of the various parts of the country. For the whole of India the ratio is 184 persons. But, as has been already ' put forward, the mean of so many and diverse component parts is of little or no value except as a standard by which the difiereuce of the said parts from each other can be judged. We may compare, for example, the above figure with the corresponding ratio in the case of France, which is 188 ; but the area of France is but about 204,000 square miles, against the 1,560,000 of India, and the population of the former is less than that of the Indian province which stands only third in this respect. To go no further than the general statement just quoted, it will be seen that this mean of 184 is exceeded in only 31 per cent, of the area, and that it is not reached, therefore, in the remaining 69. But the above 31 per cent, contains no less than 62 per cent, of the population, and when the distribution is made in smaller divisions, a rather higher proportion will be found in this category. For instance, in Statement A. below, which is compiled by districts and States, to the exclusion of the four cities coming under separate heads in the returns, the population in. higher ratio of density than the average is given as nearly 67 per cent. A. — Specific Population of India. — British and Feudatory. Percentage of Variation Area in Square Miles. Population (1891). Mean Density of each Group. Percentage of each Group on total of trom tne mean aeDsicy (,io*;. Area. Population. '75 and over - 329,650 8,246,950 25 21-13 2-89 66 50,604 2,661,892 53 3-24 0-94 rj 50 183,277 13,644,080 74 11-75 4-79 1 a 33 134,918 14,692,599 109 8-65 5-16 o 4) 20 - - 248,392 33,728,984 1 35 15-92 11-85 w 10 - - 92,654 14,580,524 157 5-94 5-12 1 ■ ' jl - - - 37,934 6,532,666 172 2-43 2-29 1^ Total below Mean 1,077,429 94,087,695 87 69 06 33 04 'I - - 50,406 9,631,878 191 3-23 3-38 10 - - - - 22,631 4,852,632 214 1-45 1-70 20 - - - - 57,682 13,179,523 229 3-70 4-63 33 47,320 12,300,679 259 3-03 4*32 i 50 36,956 10,737,381 290 2-37 3-77 a 66 8,028 2,557,1190 318 0-52 0-90 < 75, - 23,195 8,058,854 347 1-49 2-83 100 71,861 29,435,947 410 4-61 10-34 150 - ... 80,022 40,714,414 509 5-13 14-30 200 and over - 84,430 59,217,349 701 5-41 20-79 Total above Mean 483,531 190,685,847 39S 30 94 66 96 Grand Total - - 1,559,960 i 284,773,542 184 — — 78888; J) 26 This statement is based on percentages of the mean density. The first item, therefore, represents a density per square mile less than the mean figure by 75 per cent, or more, whilst the last line but two gives the figures of area and population exceeding the mean by more than 200 per cent. The table speaks for itself, but there are a few curious coincidences in it which are worth notice. The largest item of area, for example, is equivalent, proportionally, to the largest item of population at the opposite end of the table ; and, again, the proportion of population in the former , group is almost exactly the same as that in the corresponding group above the mean. Then the ratios to the total of area and population respectively very nearly coincide in the groups within 20 per cent, of the mean on both sides, whilst the divergence is, of course, widest at the ends of the scale, in both directions. There is, however, a defect in this statement which must receive mention, as it affects the regularity of the distribufc]on' between the groups in the lower densities. It is due to the fact that in the case of Kashmer and Central India there are no measurements available of the subdivisions by which the returns of the census results have been compiled. Kashmer is said to contain an area of 80,900 miles in the aggregate, but it is well known that the density is much higher in the valley and the Wazirats of Jammu than it is in the highlands of Ladakh, Baltistan, or Grhilgit. By having to enter B. — Specific Population of Beitish India.* fl Percentage o£ Variation Number of Districts.f Area in Square Miles. Population (1891). Mean Density of each Group. • Percentage of each Group on total of Area. Population. 75 and over 28 182,740 4,378,416 24 18-45 ; 2-00 66 ]4. 67,305 4,725,612 70 6-98 3-16 6 50 22 105,868 10,127,884 96 10-97 4-63 a* a 33 29 130,033 17,646,587 136 13-48 8-06 20 15 75,657 12,586,357 166 7-84 5-75 10 10 28,834 5,649,449 196 2-99 2-58 1 10 42,937 9,623,352 224 4-45 4-40 Total below Mean - 138 633,364 64,737,657 102 65-65 29-58 ri . . . 7 21,996 5,190,662 236 2-28 2-37 10 6 30,500 8,008,238 262 3-16 3-66 20 8 36,020 10,464,626 290 3-73 4-77 33 3 9,250 3,013,600 326 0-96 1-37 i ^ o 60 8 24,466 9,069,746 371 2-53 4-14 66 4 14,911 6,886,168 395 1-55 2-69 75 ' - 16 32,927 14,300,585 434 3-41 6-53 100 32 86,750 44,044,392 5i4 8-89 20-12 160 17 41,333 25,507,620 617 4-28 11-65 200 and over 18 34,302 28,730,161 838 3-56 13-12 Total above Mean 118 331,455 154,005,786 ' i 464 34-35 70-42 i Grand Total - 246* i 1 964,819 218,943,343 1 230 * ^— ^ to th SxiDluding four 8ea,port towns a n Madras the three Agency tra is ixtent- nd the detache cts are include d settlements, eoi d in their parent uprising 113 square mi district respectively, th es and 2,22 us reducing 9,509 inhabita the total num Qts. aer of districts 27 tte wliole area under its general density of only 31 to the mile, the lowest group in the return is overweighted to a considerable extent. The case is the same with the group showing a density of between 20 and 33 per cent, below the mean, because in it is included the whole of the Central India Agency, in which there are several subordinate political divisions each known to present a different density, but the areas are not accurately distinguished. Table A. can now be left, and consideration given to that on the opposite page, British marked B,, which relates to British territory alone. Here we are dealing with district terntor7. units, so the data are on a uniform basis. In the introduction it was explained that the States included most of the more thinly populated tracts of the country, so it is not surprising that their average density in the aggregate should be no more than 111, whilst in the British provinces it is 230. In Table B., therefore, the mean taken is this last figure, but in other respects the statement is framed just in the same way as the other. As regards the distribution of the population and area, it will be seen that the proportion of the former below the mean is just below two-thirds instead of 69 per cent., whilst that of the population is also slightly less. It is worth note, too, that about 45 per cent, of the population is massed in a density of 460 and over, which IB double the mean ratio of the British territory, as a whole. In Europe we find three countries with a density of 500 per mile and over ; England and "Wales, with 500, Belgium witli 540, and Saxony with nearly 600. The aggregate population of these three is about 38^ millions. In British India there are 37^ millions of the inhabitants living in the ratio of more than one per acre, or 784, on the average, to the mile. No less than 21^ millions are packed to the extent of 877. The whole of the 37^ millions in question are to be found in the Grangetic valley and the Delta, and with the exception of 2i millions, ^hich may be called suburban, live, in the main, by cultiva- tion of the soil. It will be necessary to revert to these tracts in the next chapter, when the increase or decrease of the population is in question. If we refer to the other end of the statement, we find over 182,000 square miles with an average population of 24 each. The greater part of this group falls within Upper Burma, the hill tracts of Assam, and the desert portion of Sindh. With regard to the general distribution shown in the rest of the table, it will suffice to note, first, how very small a proportion of either area or population lies within 10 per cent, of the general mean, whether we take it above or below that figure. This fact bears out in a striking- manner the statement made above, that the latter cannot be used in discussing the census results as typical of the country, but only as a means of measuring divergencies in various parts of it. Again, as to the composition of the larger groups below the ' mean, it is worth mentioning that nearly half the Central Provinces and Berar falls within the group averaging a density of 136 per mile. To the next group, with 166, the northern districts of the Madras Presidency, and the Bengal portion of the Central Belt, separating the Ganges basin from that of the Mahanadi, are the chief con- tributors. The groups just' noted above as lying immediately round the mean, with an average density of 224 and 236 respectively, include the remaining districts of the North Coromandel coast, and a considerable portion of Bombay. Then, crossing the line, we come to the districts of the southern plain and of parts of the Panjab. Passing over the next few groups, which are formed out of a few tracts from several provinces, we come to that of which the mean density is 434 per mile, or nearly double the general average. The bulk of this is contributed by the west-central tract of the North- Western Provinces, with the aid of a few sub-montane districts of the Panjab, and of the less thickly peopled tracts of Bengal. "When we reach the higher specific rates, we find only a single district of Bombay, a couple of Madras, and five of the Panjab, standing out amidst the enormous masses of the Gangetic valley. Before entering further into the details of the provincial distribution of the popula- tion according to density, it will be interesting to deal with the subject in its more general form, in connection, that is, with the distribution of both area and population according to rainfall. For this purpose,, Table C. is appended. It includes the whole of India, British Density and Feudatory, for which rainfall returns are available to a trustworthy extent. For ^°^P^^?^, reasons already given in the introduction, those for Lower Burma are too partial for ^|]_ use, and it would be misleading to enter the whole of such widely spread tracts as Central India and Kashmer without splitting them up according to_ the known variety of their conditions as to humidity. The detached Settlements, again, are too small to be taken into consideration, even if full information respecting them were at hand. The omissions, being mostly those of sparsely populated country, have raised the general D 2 28 average of density to nearly the mean between that of India, as a whole, and that of British territory alone, as given in the two preceding tables. C. — Mean Eainfall of India. Composition of rainfall — groups. Eelation between rainfall and density. Percentage of each Group on total of , Mean Percentage of Variation Area in Square Miles. Population (1891). Mean Density. Rainfall in from the mean Eainfall (42"). inches. Area. Population. ^75 and over 123,394 6,708,577 54 9-40 2-49 7 13 66 66,826 4,429,726 66 5-09 1-64 S 50 - - - 35.708 5,658,222 158 2-72 2-10 17 s 33 165,366 28,262,697 171 12-59 10-48 25 -r <; 20 - 184,867 36,780,747 199 14-07 13 62 30 o 10 89,824 22,171,842 247 6-84 8-22 35 1 142,898 38,102,505 267 10-88 14-12 41 Total below the Mean 808,883 142,114,316 176 61-59 52-67 26 ' 1 85,016 25,878,282 304 6-47 9-59 46 10 72,810 22,607,267 310 5-54 8-38 49 20 79.970 12,832,274 160 6-08 4-76 53 s 33 • - 81,003 18,290,630 226 6-17 6-78 60 C3 ID 50 44,196 12,877,074 291 2-37 4-78 68 ^ 66 2,797 2,420,656 865 0-21 0-90 73 75 42,126 6,590,947 156 3-21 2-44 77 o 100 42,444 10,584,251 249 3-23 3-92 91 ^ 150 39,001 10,963,712 281 2-97 4-06 115 200 and over 15,168 4,643,949 306 1-16 1-72 140 Total above the Mean 504,530 127,689,042 253 38-41 47-33 66 Total 1,313,413* 269,803,358* 205* — 42* * Excluding Kashmer, Central India, Lower Burma, and the detached Settlements. The first point to note is the large proportion of the population found in tracts with less than the mean rainfall. In connection with this there should be borne in mind what was said before, as to the relative position of some tracts with regard to the certainty of their seasons, irrespective of the amount of rain actually received. For example, in the largest group in point of population, that having an average of 41 inches of rain, we find the Hill States of "Western Bengal, the submontane tracts of the North-West Provinces, parts of Oudh and Bihar, and the northern portion of the west coast, all of which enjoy comparative certainty of seasons, mixed up with others that are far less fortunate, such as parts of Haidrabad and Madras, in the Decci&n, and the outskirts of the uncertain zone in the Ganges valley. On the other hand, its neighbour, just above the average, which is not nearly so populous, may be said to lie almost entirely within the lucky line of the rainfall, and contains much of the hilly tract in the Central Provinces and the eastern portions of the submontane tract of the North-West, with a good deal of the best parts of, Oudh. Going back to the rainfall of about 30 inches, where the largest area in any single group is to be found, wo may see that the few entries from the Panjab, Berar, and Bombay are outweighed by the number from the Deccan and the worst portions of the North-west Provinces. The most ill-situated tracts of the Deccan, however, are found in the two groups preceding this one. The groups of lowest rainfall, those with no more than 7 inches and 13 inches respectively, we find situated in a part of the country where the aid of rain is, as a rule, discounted by the extension of irrigation from rivers fed by Himalayan glaciers, or from deep wells, or, again, by watercourses led down from hill streams. A second matter to be noted in connection with this table is the sudden break in the series of increasing densities of population which occurs where the rainfall advances from 49 inches to 53. Below this, in the series, there is a continual, though irregular, rise in density from the lowest group, but the same feature does not occur again until within the three last. The explanation of the break, in the first instance is that the group in question begins to touch the hill country of the Central Belt at its eastern extremity, where the fall of rain is comparatively heavy, but the arable area is scattered amongst forests, so that in no parts is it possible for large congregations of people to find sustenance. In the next group, the remainder of this tract has its lowering effect on the average outweighed by the inclusion of the less rainy portions of 29 Bengal, where the density is far greater. In the next few classes the last-mentioned province is mixed with the thinly populated Assam, and the coast tracts of Bombay, where they are least .fertile, have also to be taken into account. The Malabar States and districts raise the densiljy again in the last but one of the groups, whilst that of the highest rainfall comprises several of the Bengal submontane districts, the most populous district of Assam, and one of the corresponding class in "Western Madras. Had the figures for Lower Burma been included, the group-density would have fallen to 88 per mile, instead of being 306, but the average rainfall would have risen to 170, a result which, as has been already stated, is not in accordance with the circumstances of more than the littoral portions of the tract. The population has been considered hitherto in its distribution over the country as a whole, or at least in its two great political divisions of British and Feudatory. It has now to be treated in the greater detail of Province or State. The diversity of the distribution, even in these smaller units, has been indicated in the remarks made above in explanation of the general statements, and is illustrated in a still more striking manner in Statement D., given below, by the addition of exponential figures to the density of all provinces that contain a considerable number of districts. It has already been explained that in the case of States, or groups of States,- such additions are useless. D. — Specific Population op Provinces and States. Mean Density per Square Mile. « Extreme Density. Province or State, &c. Highest. Lowest. Name. Density. Name. Density. Ajmer - Assam 200 68-45 112 43-99 471 2S-8o 163 35-71 207 48-33 60 Ajmer - - - Silh^t 203 397 Merwara - . - Kaohar Hills - 187 11 Bengal Berar Bombay Saran Amraoti Kaira • , - 930* 238 542 Chittagong Hills Elichpur Kanara 20 120 114 Sindh Haidrabad 102 Thar and Parkar 23 ■ Upper Burma 117 -14 •35 92- 15 53 29-60 125 Mandalay 178 Upper Khjndwin 4 . Dower Burma Central Provinces Tharawaddi Ifagpiir 172 197 Salwin Chanda 7 65 Coorg Madras 109 49-01 252 Tanjore 601 Godavari Agency 42 "N.-W. Provinces .Oudh 42-09 411 21-68 522 Benares Lucknow 914 800 Garhwal Kheri 72 305 Panjab Total, British 79-78 18S 55-65 230 Jalandhar SAran 633 930* Dehra Ismail Khan Upper Khyndwin 51 4 Haidrabad - 139 Bidar 216 Sirpur Tandvir 46 Baroda 294 Baroda 428 Amreli 151 Mysore 177 Bangalore 314 Chitaldurg ■104 Kashmer 31 Eanbir Singhpiira — Ladakh — Kajputana 92 Bharatpur 327 JesalmSr - 7 Central India 133 Saghelkhand-f 1.53 BUndilkhand 72 Bombay States 126 Kolhapur 334 Sdrgdna 34 Khairpur QSindh) 22 — Madras States 385 Cochin 531 Sandur 71 Central Province States 73 Chhuikadan 235 Bastar ?4 Bengal States 92 K6chh Bihar 443 Hill Tippera 33 N.-W. Province States 155 KSmp6r 583 Tehri-Garhw41 - 58 Panjab States Total, States 111 60-81 111 Kapurthala _^ X- — 501 Bashahr 23 * Excluding the suburban district of Howrah. D 3 f The Central India return is approximate only. 30 Haidrabad. Baroda. Mysore. Kasmer, Kajputana, Central India. In the above table, the density, taken item by item, appears to vary little more in the Feudatory than in the British territory, and in both, the extremes of crowding and sparsity of population are found very far apart. As the Provinces can be advantageously treated in greater detail, it will be convenient to deal with the States first. To begin with, the larger ones, Haidrabad, Baroda, and Mysore, are all sub- divided into something corresponding to the district in British territory, so the distribution can be fairly appreciated. The first-named shows the highest density in the portion adjoining the uncertain zone of the Deccan, but as it lies more to the east, it gets the benefit of some of the autumnal rains. The lowest density is found in the north-east corner, at the edge of the southern extension of the Central Belt of hills, where it bounds the Central Provinces. In the rest of the State the density is fairly uniform. Baroda has the densest population of the larger States, if we except the two that round oif the Malabar coast, but, as the table shows, there is considerable variety in the distribution. The greater portion of the State lies embedded in the British division of Gujarath, the most densely peopled tract in the Bombay Presidency. In the east of the State there is a large forest area, which reduces the general average, whilst a portion of the State lies in the peninsula of Kathiawar, and is here shown as containing, relatively, the lightest population. In Mysore we find the heavier densities are in the south part of the State, round about the two large towns. The north-eastern tract, showing the lowest, borders on one of the least favoiired districts of the Madras Presidency, Belari, and shares the physical characteristics of the latter. In dealing with Kashmer, nothing more can be said at present beyond mentioning the tracts which are reported to contain the heaviest and the lightest densities respectively. Probably the population of the portion of the valley lying immediately round the chief city, Shrinagar, is relatively as heavy as that entered in the statement, and, on the other hand, the (jrhilgit region is as lightly peopled as the eastern table-lands. Eajputana contains at least three well-defined regions, with very dififerent specific populations. In the east, where the States touch the G-angetic provinces, the density is over 300 per mile in one State, and over 200 in the two adjoining it. Central Piajput^na, which is combined with the eastern part in some of the returns, varies between a density of lfi4 in the largest State, Jaipur, and of 113 in Jhalawar, where the country trends upwards from the plains to the plateau of Central India. The southern portion is different, again, in the direction of less dense population. The only large State in this division is Mewar, or Udaipur, the premier Eajput State in India, which has a density of 145. Its neighbours are mostly small offshoots of families of chieftains who have been more fortunate in their settlements, so that the Vounger branches have to content themselves with the allegiance of the hill tribes of this portion of the Central Belt, who number little more than 100 per mile. Lastly, we find the sandy plains of the west. The leading State of this tract is Marwar or Jodhpur; with its 67 inhabitants to the mile, flourishing under 13 inches of uncertain rain. Bikaner, its neighbour to the north, receives an inch less, and supports but 36 to the mile. To conclude, comes Jesalmer, with 7 inches of rain, 16,000 square miles of teipritory, and but seven inhabitants per nlile. Yet this western region, desert, we may call it, not only breeds the finest camels in India, which might be expected of it but sends forth to every important market in the country the keenest and the most frugal Set of traders that ever drove a bargain. It is probably the large fortunes that are made by these men, and the animus revertendi alniost invaribly maintained by them during their term of spoiling the Egyptians, that have kept up the general well-beino- of this tract, and the undoubted wealth of the three or four large towns that it contaiiis.* Passing south, to Central India, we find an increased density, on the whole but the details of the different parts of this large agency cannot be relied on, so far 'as areas are concerned. The States are rarely compact, but consist of many scattered patches of territory, one within the other. But within broad lines, the two selected for the table may be considered approximately correct. The parts of Rewa that border the Ganges valley contain a population grouped in considerable density, whilst the rest of the State consists of hilly country of much the same class as that in the Biinddlkhand * There is iilso the tradition of safely, as these tracts have been generally avoided by Muse! "W ' and other marauders, ONving to the absence of supplies for any considerable force, and the waATke VZu.% the local chieftains. vm.nijj.i, lepuie ot 31 oliarge, wliere tlie density is probably the lowest in Central India, save, perliaps, tbat in a few small tracts intbe extreme west, amongst the Vindhya hills. The States in political connection with the Bombay Presidency are, for the most Bombay part, small, like those under the general title of Central India. But, unlike the latter, States, their territory is mostly compact and does not intermingle to any considerable extent. For the present purpose, the densities have been calculated according to the groups under distinct Agencies, as many of the States are too small to require individual recognition in the returns. The relic of the Talpur dynasty in Sindh, found in Khairpur, has been separated from the rest, and shows a specific population even less than that of its neighbour, Bahdwalpur, across the frontier of the Panjab. Through- out the Presidency proper, the density of the Feudatory territory is greater. The lowest is in the forest States, where the Satpuras and Grhats subside into the plain bordering on the coast. Further north, in Gujar^th, there are wide expanses of light sandy soil, as in Kachh and Palanpur, and hilly country, such as is found in most parts of the groups along the B,ewa (Narmada), and the Mahi rivers, and in the Surat Agency and its neighbours. The population thickens along the coast, falls in the Deccan, and attains its greatest density in the South Deccan, or Bombay Karndtak, of which Kolhapur is the leading State, with 324 persons to the mile. The States in connection with the Madras Presidency are bat five in number, and Madras of these two are very small. They are both situated within the zone of uncertain States. rainfall, and have a population of but 71 and 139 per mile respectively. On the Ivlalabar coast, the circumstances are more favourable to life, and Cochin comes high Up in the scale of density. Its larger neighbour, Travancore, though retui'ning but 380 to the mile, is really more densely populated, since behind the lowlands along the coast there is a considerable extent of hill country almost uninhabited. The States connected with the Central Provinces are, in most cases, little more Central than private estates, to which has been given the status of political units within Pioviuce comparatively recent times. Of those in the extreme east of the province, the leading ^*''^'^^- State of years gone by is still the most populous, though the relative incidence is smaller than that found in some of the others, which have profited by the opening out of the neighbouring tracts by roads and rail. The largest State, Bastar, consists chiefly of hill and forest, with so widespread a reputation for fever and bad water that it has been left to the dark tribes, to whom alone such tracts are congenial. A few States in the centre of the province, which include within their limits a little of the plains, show a higher density, reaching in parts to 200 persons per mile and over, but for the most part, the dominion of these chiefs lies in the hill and forest, where population is necessarily sparse. The circumstances of the Orissa States, of which there are 17 in political con- Bengal nection with the Government of Bengal, are almost the same as those of the eastern States. States under the Central Provinces, and show a density pf 118 per mile, varying between 43 and 447, though the last figure is reached in a minute estate of little over 40 square miles in extent. The Chiitia Nagpur States, which spread to the north- westwards of the last group, are still more thinly peopled, and. show an average of only 55 to the mile. All of these are more or less within the domain of the dark tribes, whose mode of life,, with its system of exploiting the soil, is i exactly that of the " rigidi Getae " of Horace : — " Immetata quibus jugera liberas Fruges et cererem f erunt ; Nee cultura placet longlor annua." A similar state of things will be found in the small State of Hill Tippera, between Bengal and Burma. If anything, it is worse here, because, owing to the universal habit of cultivating by the primitive mode of clearing a plot of forest and burning the timber and brushwood off it as manure for the one year's crop, the local resources are soon exhausted, and if virgin soil is not at hand the population has to flit a few miles to places where the forest has had 8 or lO years rest. Thus there is little in the shape of a permanent village in the State, the Only place of this character being the residence of the Chief. Lastly, we have the State of Koch Bihar, on the Assam frontier, where the land has long been more or less brought under cultiva- tion. Here, the density of the population has risen to 443 per mile, which is within about 20 per cent, of that of the adjacent British territory, and much above that of the eastern frontier, which divides it from its parent region, Assam. , There are only two States politically in connection with the North-West Province N.-W. Government, and these two are widely separated as regards population, as can be seen Province T) A « . •• , States. 32 Pan jab States. Comparative densities of British Pro- vinces. Use of the weighted mean. The smaller Provinces. from the table under consideration. Eampur is embedded in Robilkhand, and the Chief is one of the tribe or tribes from which that tract takes its name. The other State lies in the Himalaya, and is little but a succession of hills and dales, from the outer slopes to the highest peaks of the group in which the Ganges and Jamna have their sources. In the one the density is that of the surrounding portions of the rich Granges valley, in the other no more than 58 per mile. The States connected with the Pan jab Grovernment may be divided into four main groups. First, those in the eastern plains, which appertain to the Gangetic system. Here the density is about 263 per mile, rising in Nabha to just over 300. This is the area formerly known in history as the Cis-Satlaj Protected Sikh Stafes. Then come the States of the submontane tract, the most densely peopled of the whole body, though with a total population of less than one-third that of Patiala, the largest of the group just mentioned. There is one State, large in area, in. the western plain, bordering on Sindh and the Rajputana Desert. Bahawalpur is a little lai'ger than Jesalm^r, one of its neighbours, but contains nearly six times the population. Even then its density is but 38. Last of all come the Hill group of States. Most of these lie amongst the loWer ranges of the Himalaya, but two run up to the central mass of these mountains. In these last we find, of course, the lowest densities, as so much of their territory is little more than mere rock and snow. The population of Bashahr, for instance, is but 23 per mile ; in Chamba, too, it only reaches 40, and for the whole group, the average is 7?. On looking at the Return B. for British provinces, to which attention is now to be directed, we find a maximum divergence of more than 700 per mile from the average, for in the Upper Khyndwin district of Burma, which, however, has not been completely surveyed, there seem to be only*four inhabitants to the squai'e mile, whilst in Saran, a district in the Bihar division of Bengal, there are 930. It is not easy to find corresponding densities to those of many of these provinces amongst the countries of Europe, but the few that are fairly comparable are as follows. Tn the Panjab the specific population is about that of France. Madras compares well with Italy, and Bombay with Austria. Hungary represents the Central Pro\';inces. Berar comes nearest to Denmark, and Assam to Servia. The heaviest density of a province, as a whole, is found in Oudh, where it is about 20 persons per mile below that of Belgium, the second in rank in this respect of the European countries. After this we must look to fractions or multiples for our analogies. Sindh, for instance, shows double the density of Sweden, and Bengal exceeds in like proportion the density of the German Empire. In Burma, as a whole, there are about half as many people per mile as in Spain, and so on. But the exponential figures in the second column of the statement, taken -.with the extremes given in the succeeding part of that table, show how little uniformity there is in the density of the various parts, even of the same province. There is none in the list where the average divergence is less than one-fifth. Oudh comes first in this respect, then Berar, which has but six units to consider. The Central Provinces run it close, however, in spite of the great divergence of the extremes. But when we come to the Panjab and Assam, still more to Burma, it is plain than the mean is no more than an arithmetical expression, of use only for comparison af the whole province with its neighbours. The distribution, therefore, of each province, by districts, on the same lines as have been adopted in Table B. above, is shown 'in Statement B. on the opposite page. In consideration, moreover, of the variety of the larger provinces in point of density and rainfall, the natural groups into which each province falls will be found specified marginally, as the province comes under review below. The reader should also note that in this review the density, and not the pressure,, of the population on the land is in question. It is convenient to take the less densely populated provinces first in order especially as several of them are not subdivided. Coorg, for instance, is shown in one item in the tables, so is excluded from Statement B. The greater portion of this little tract consists of forest and hill, with clearings for cultivation interspersed. Much of the population, too, is migratory, and only inhabits Coorg during the season when the coftee and other plantations offer special attractions to field labour. Lyinff along the bhats, Coorg enjoys a heavy and fairly certain rainfall, and bears with ease its hght population. Ajm^r, again, consists of only two divisions, Merwara, and that which gives its name to the tract. The latter is traversed by two lines of rail and oontainfi two towns -t»f- considerable population, to which some of its density is ' due 33 E. — Peovincial Density G-roups. ASSAM. panjAb. Density per Square Mile. No. of Districts. Area. Fopulatioo, 1891, Density per Square l^e. No. of Districts. Area. Population, 1891. 57 and below 5 20,249 504,916 ' 57 and below 1 9,440 486,201 58-75 1 3,724 254,053 58-75 3 14,248 1,044,047 i 76-114 115-152 3 1 10,630 2,472 1,104,206 367,542 1 76-114 115-152 6 2 33,516 9,158 3,140,944 1,385,062 ■3 153-183 2 6,515 , 1,091,523 1 153-183 2 7,835 1,403,482 184-206 — — — 184-206 1 4,302 886,676 _307-229 — — - _207-229 1 3,017 690,169 '230-253 — — — '230-253 — — 254-276 — — - 254-276 — 277-307 — — — 277-307 3 8,562 2,462,865 . 308-345 — — a 308-345 2 3,781' 1,269,404 1 346-384 — — — 1 346-384 2 4,805 1,794,302 1 < 385-402 403-460 1 5,414 2,154,593 o < 385-402 403-460 3 3,799 1,705,023 461-575 — - — 461-575 3 5,170 2,702,458 576-690 — — — 576-690 2 3,034 1,900,280 691 and over — — — ^691 and over — — - Total 13 49,004 5,476,833 Total 31 110,667 20,860,913 Baloch tribes — — 5,934 I 78388. BOMBAY. MADRAS. Density per Square Mile. No. of Districts. Area. Population, 1891. Density per Square Mile. No. of Districts. Area. ■ Population. 1891. 57 and below — — — 57 and below — — — 58-75 — — — 58-75 — , — s 76-114 1 3,910 446,351 i 76-114 2 8,471 917,608 115-152 4 29,160 3,989,527 115-152 3 19,972 2,880,747 153-183 1 4,542 750,689 ^ n 153-183 2 26,007 4,266,728 184-206 2 6,982 1,381,217 184-206 — — — 207-229 2 9,260 2,064,575 207-229 2 16,767 3,752,385 ''230-253 4 14,334 3,394,059 '230-253 — — — 254-276 1 1,872 509,584 254-276 4 27,148 7,102,293 277-307 1 3,922 1,105,926 277-307 2 16,424 4,788,891 308-345 — — — d 308-345 — — — 346-384 — — — 1 346-384 2 9,018 3,288,812 385-402 1 1,662 649,989 01 385-402 1 2,842 1,136,928 ^ 403-460 — — — ^ 403-460 1 5,217 2,162,851 461-575 f 1 1,609 871,589 461-575 1 5,585 2,652,565 576-690 — — — 576-690 1 3,709 2,228,114 691 and over - — — — 691 and over - — — — Total - 18 77,263 15,163,606 Total - 21* 141,160 36,177*922 Bombay City - ~ 22 821,764 Madras City - — ; - ^' 4Sz,Si8 * The three Agency tritcts are included in their parent districts. E 34^ E . — Provincial Density ^nov^s—cmctuded. N.-W. PROVINCES AND OUDH. BENGAL. Densitj' per Square Mile. No. of Districts. Area. Population, 1891. Density per Square Mile. No. of Districts. Area. Population, 1891. ' 57 and below — — — ' 57 and below 1 5,419 107,286 58-75 1 5,629 407,818 1 58-75 — — — ' i 76-114 1 7,151 563,181 [ § 76-114 1 1,681 170,058 ^ 1 115-152 2 3,140 442,335 115-152 2 8,658 1,142,258 153-183 — — — 153-183 * 2 14,161 2,293,206 184-206 — — — 184-206 1 1,164 223,314 .207-229 3 8,475' 1,885,796 .207-229 — — — '230-253 2 4,700 1,115,251 '230-253 1 2,962 681,352 254-276 1 1,480 396,361 254-276 — — — 277-307 :- 1 2,965 903,615 277-307 1 4,147 1,193,328 308-345 — — — . 308-345 1 5,469 1,754,196 1 346-384 2 4,052 1,485,798 M' 346-384 2 6,591 2,500,833 1 385-402 ■— — — t 385-402 1 4,993 1,944,658 ^ 403-460 7 12,923 5,611,960 '<. 403-460 4 10,988 4,820,751 461-575 14 29,006 14,799,734 461-575 13 44,379 23,018,046 576-690 8 17,947 11,366,223 576-690 6 16,643 10,013,003 ,691 and over - 7 10,035 7,927,013 .691 and over 11 24,268 20,803,138 Total 49 107,503 46,905,085 Total 47 151,523 70,665,427 Calcutta — zo 68i,56o BERlR. CENTRAL PROVINCES. Density pejr Square Mile. No. of Districts. Area. Population, 1891. Density per Square Mile. No. of Districts. Area. Population, 1891. 1 1 Ra ' 57 and below 58-75 76-114 115-152 153-183 184-206 207-229 Total ilwai/ Passengers 3 1 2 6 9,'490 2,809 5,419 17,718 1,185,592 481,021 1,230,427 2,897,040 45 1 1 o -33 m ' 57 and below 58-75 76-114 115-152 153-183 184-206 .207-229 Total 3 2 7 2 4 18 19,162 8,454 37,834 7,376 13,675 86,501 1,290,469 730,690 4,949,984 1,197,267 2,615,884 10,784,2i94 SINDH. -■ ■ __ 1 BURMA. Density per Square Mile. No. of Districts. Area. Population, 1891. Density per < Square Mile, i No. of Districts. Area. Population, 1891. i 1 '' 57 and below 58-75 76-114 115-152 1.53-183 184-206 ^207-229 Total - 2 1 2 ; 5 26,911 2,549 18,329 47,789 863,083 174,548 1,834,143 2,871,774 1 1 i ^ 57 and below 58-75 76-114 r 115^152 163-183 184-206 .207-229 , Total Rangoon Town 19 5 3 5 3 35 120,721 21,993 12,133 10,149 • 6,412 171,408 zz 2,416,930 1,554,677 1,047,648 1 1,303,540 1,102,441 7,425,236 i8i,3z4 .'-.iii'ijsri .'•. 35 Berar. as the surrounding tracts are not, on the whole, fertile. The smaller tract, of Merwara owes its present comparative prosperity to a single British officer, Colonel Dixon, whose personal influence during his long residence there served to reclaim the wild tribes and to establish a thriving centre of trade for the whole country round. As regards density, the table shows that in the present day, at all events, there is not much difference between the two districts. The detached Settlements of Quettah, Aden, and the Andamans require a few words here. The area is only ascertained in Aden, &c. the case of Aden, and for that only approximately. This Settlement consists of three small tracts. First, the peninsula, on which the main population resides. Secondly, the suburb on the adjacent mainland, including Shaikh Ofchman, and the Somdli quarters. Thirdly, the Island of Perim, about 90 miles up the Straits of Babelmandeb, on which there is a small detachment of military, and, of late years, a coaling depot. The census of Quettah included only the cantonment known by that name, with the Quettah. adjoining town and the railway settlements, together with the outlying cantonment of Lorelai, and the camps of the political officers, who were mostly collected at the time of the census at Sibi, below the Boldn, for a horse fair. Similarly, the census of the Andamans was limited to Port Blair, so we have no record of the natives, who are Andamans said to be dying out in some of the islands. We then come to the five districts of Sindh, the most dense of which has only 102 persons to the mile. The lowest, Thar Sindh, and Parkar, lies, as its name implies, on the outskirts of the desert, and supports its 23 to the mile chiefly on the charitable relics of the south-western rains, received up the Gulf of Oambay. This province presents no such striking differences within its limits as call for subdivision of the area, though if the tables admitted of more minute partition, no doubt it would be practicable to separate, the deltaic portion, where there is a light rainfall of about 8 inches, from the upper regions, where cultivation depends upon the inundation or canal irrigation. Berar, again, is taken as homogeneous in its features, though there are small hilly tracts in both the north and south-east. The greater part lies in the Central Plain, and the density at present does not, in any tract, reach the mean of British India. The rainfall, though light, 354 inches, is fairly steady, and the soil famed for its growth of cotton and oil -seeds as well as of millet. Its neighbour on the east, the Central Provinces, is the first in which subdivision is necessary. The mean density of 125 per mile is composed of three main factors. First, the fertile rice-plains ©f the east ; then the valley of the Narmada, and lastly, the belt of hills that spreads west- wards from the Chutia-Nagpur plateau. North, east, and south, is found a fringe of lightly peopled forest and hilly country, where the" absence of large areas of flat land prevent the formation of populous villages, in spite of the comparatively certain rainfall. Burma is the province to be reviewed next, and it is the last in which no district Burma, is found to have a population of the mean Indian density. Broadly speaking, it may be divided into no more than four tracts, two, littoral and deltaiic, in Lower Burma, and two, central and submontane, in the upper portion of the province. But it has been found of practical con- venience to group the districts according to the administrative divisions, as shown, in the margin. Here, as has already been remarked, the mean figures help us not at all. No one of the districts comes within 30 per cent, of the general average, and the range is from 4 on the Khyindwin, to 178 in the district containing the capital of Upper Burma, and from 7, in the southern coast hills, to 172 in an older settlement of the lower portion of the province. By far the greater number of the districts lie in the two lowest groups of density, with less than 75 persons to the mile. That this light population is due to the actual want of men, not to the inability to support them, is proved by the very large proportion borne by the amount of rice exported to that grown, though it is the staple food of the whole province. E 2 Tract. Dis- tricts. Area. Density. Rainfall. Eastern Southern Central J^.n-rmada 3 4 5 6 25,013 20,988 19,847 20,653 142 124 91 136 53'-'49 54-56 50-91 46-67 Central Provinces. Tract. Dis- tricts. Area. Density. Eainfall. . ["Northern S J Central a 1 Southern ^ [Eastern . r Arakan 1 J Pegu 1 Irrawaddi ^ L Tenasserin 5 4 4 4 4 5 4 6 22,580 29,236 18,607 13,050 14,526 9,299 17,542 46,590 28 22 42 59 46 156 88 21 47'-00 56-48 25-13 40-75 191-11 65-03 84-01 160-04 Traijt. Dis- tricts. Area. Density. Rainfall. Siirma Valley Brahmaputra Valley Hill Tracts 2 6 5 7,886 20,869 20,249 333 117 25 146'-'27 87-12 92-03 Assam. Assam, which comes next on the list in order of sparsity of population, is really/ m the same category as the preceding pro- vinces, save as regards the one district that partakes of the character of those in Bengal that adjoin it. Twelve of the 13 districts contain an average of little more density than Sindh, or about 76 per mile, whilst Silhet shows 398. "We may separate, accordingly, this with its neighbour Kachar, from the rest, under the title of the Surma valley. Assam, properly so called, consists of two divisions, the BrdhmapuT-ra valley, and the Hill tracts, five in number. The latter fall into the lowest group as regards density, and have a specific population varying between 37 in the Garo Hills and 11 in North Kachar. In the Brahmaputra valley the extension of tea plantation, apart from the general opening up of the country, has been the means of attracting a considerable number of immigrants, chiefly from^ Bengal, many of whom settle on land- grants made to them after their term of service on the tea- gardens is completed. It is to be noted, too, that the eastern portion of this valley has been a frontier tract for centuries, so that till within recent times the population was kept down by war and invasion. This explains the greater density of the western districts. The river, too, has been much more used of late years than before, as the main line of communication between Assam and other parts of India besides Bengal, but the prevalence of an epidemic of fever, which has been peculiarly fatal in this valley, seems to have checked the growth of the province, as will be seen from the statistics given in the next chapter. Panjab. Missing the main portion of the Gangetic basin for the present, we come to the PanjJib. The figures in Table D. alone show how necessary it is to subdivide a tract that exhibits so wide a divergence between the extremes of its component parts. On the other hand, Table E. shows that this province, as regards the number of its districts, lies nearly equally above and below the Indian mean, and much of it enjoys a nearly uniform rainfall. The five divisions selected, however, show where the differences are found. First, we have a submontane belt in which the density is in places nearly equal to that of the Gangetic valley, to which the eastern portion of the province may be said to belong, though the proximity of the sandy plain to the south keeps down its density. The next division is scarcely so well defined physically as racially, but is well enough described as the " Salt Bange Tract," wherein the density sinks to less than half that of the submontane. The two other divisions are plainly denoted. The western plains are not in great part the desert that they seem to be from the map, or eve;a from the small density indicated in the margin. The soil is fertile enough, but requires water to render it available for cultivation, and every decade sees irrigation canals more and more extended from the snow-fed rivers which join the Indus where it is leaving the province for Sindh. The hill tracts, which are the last to be mentioned, consist largely of the great stretches of Himalayan rock and snow, such as have been already described as characteristic of the States in these parts of the country. Bombay. The next province that engages our attention is that of Bombay, and here, again, we find that out of the four main di- vis'ons, three are based on physical con- dit( )ns, whilst in regard to the fourth, ra( ■ differences have been also taken iiito account. The first division is the expan- sion of the coast line, after the Ghats have receded into the interior. The second is the strip of coast itself. The third is, of course, the Deccan plateau, so far as it lies within the limits of the in. western Presidency, whilst the last is the southern part ot that plateau, bordermg on Mysore and north Madras Koughly Tract. Dis- tricts. Area. Density. Rainfall. Submontane and Central. Eastern Plains Salt Eange Tracts Western Plains Hill Tracts 11 5 6 7 2 20,413 12,674 20,892 41,012 9,676 381 265 158 81 83 ! 26'-'79 25-70 25-31 7-50 54-97 Tract. Dis- tricts. Area. Density. Rainfall. Gujarath 5 10,296 301 29'-'44 Konkan 4 13,639 210 104-66 Deccan 6 38,390 162 30-36 Karnatak 3 14,928 192 34-15 Bombay Island - — 22 — 74-23 37 Tract. Dis- tricts. Area. Density. Eainfall. I^orth Coast 5 31,464 279 48'-'29 Northern 4 27,486 134 25 47 East Central 3* 10,487 369 39 51 West Central 3 16,346 249 30 45 South Central 3 12,5S7 458 44 48 Southern 2 14;195 319 32 46 West Coast 2 9,487 391 117-17 * Including Madras City. speaking, the Karnatak begins soutli of the BMma river, on the east, and in the State of Kolhapur, in the west, without other marked geographical demarcation. The city of Bombay has been excluded in the tables from the Konkan, to which it belongs, so as not to disturb, with its peculiar circumstances, the general distribution of population along the coast. Considering the number of the districts and the difference of the tracts into which they are thrown in the returns, the mean figure is remarkably in accordance with the detail. GujtoAth presents the greatest density with the lightest rainfall, owing tJhiefly to the large area of fertile soil in the northern part of the division. The Konkan, which is the local name for that part of the coast, comes next, but here the rainfall rises, and the delisity recedes, by reason of the comparatively narrow area of cultivable land. In the Deccan has been classed a large portion of the central plain, included in the district of Khandesh, and the two large towns of Poona and Sholapur also tend to raise the population. The rich black soil mentioned in the first chapter as peculiar to this tract is found in the Karnatak in wider and more continuous expanse, and in the south-western portion of this last division the rainfall is heavier and more certain than to the east and north of the Deccan proper, so that the density is considerably greater. The Madras Presidency, with which we have now to deal, has probably the greatest Madras. number of climatic differences within its limits of any, though their range is not so wide as in the Panjab. The districts are thus very widely distributed in Table B., but, with the exception of the Agency tracts on the north coast, which really belong to' the Central belt of hills, there is little land lightly peopled. The zone of uncertain rainfall embraces the M'hole of what is mar- ginally termed the Northern division, at perhaps its worst point, and owing to the protective effect of the western Ghats, the West-central division is little better off. On the other hand, along the north coast and in the South-central divisions, there are lai-gely extended systems of irrigation from the deltas of the Kistna, the Godavari, and the Kavari, by which rice is grown enough to support a very heavy population, almost irrespective of the annual rains. The western coast districts, on the contrary, depend upon the south-western current, here at its maximum of strength and humidity, so they attain the same result by different means, and their produce, if not so near the foreign market, is probably more varied and to some extent, where spices and condiments are cultivated, more valuable. Last of all, the task remains of reviewing the manifold conditions of the Gangetic valley, as comprised within the two great provinces of the north-west, with Oudh and Bengal. What has been said already of these tracts in connection with different subjects, has indicated the remarkable density of the valley as a whole. Even taking into consideration, as is done in the table under review, the outlying territory included in each of the two provinces, the specific population of Bengal rises con- siderably above, and that of the other province comes not far below, double the mean of British territory, as a whole. In the former, 76, and in the latter 72 per cent, of the population exceeds that average. Taking the two political divisions of the Upper provinces separately, it will be ]sr..w. Fro- found that in the North-wer^f^rn division vinces and the density is only 411 per mile. In Oudh. Oudh, however, it rises to 522, a ratio unparalleled over so large an area in any Indian province except Bengal. The density of the combined Pro- vinces is 436. The latter has been divided into seven physically different tracts. One of these, situated amongst or just below the Himalaya, is thinly populated by reason of the scarcity of arable land. There are parts, indeed, of this tract that are deep in snow most of the year, so that the population descends for the winter into the grazing country at B 3 Tract. Dis- tricts. Area. Density. Eainfall. Hiraalaytm Submontane Upper poab Central Dodb Ifoith Central - South Central Southern 3 5 6 7 15 6 13,973 18,202 10,133 10,139 14,116 25,300 15,639 81 486 509 470 499 652 221 62'- 13 46-43 31-25 28-38 38-29 38-42 37-51 the foot of the hills and returns only for the four or five months when cultivation is practicable. This migratory population, therefore, was added at the time of the census to that of the submontane tract, but the total number of persons thus displaced was comparatively insignificant. At the opposite side of the province, to the south, lies another tract that shows a density which, though it is not far below that of British India, as a whole, must be classified as sparse when considered in relation to the rest of the province. It consist largely of broken ground, forming the outworks of the Central belt of hills. Between these extremes, the valley rarely supports a population of less than 460 per mile. The exceptions are found in tracts along the base of the Himalaya, where, for some reason or other, there has been a good deal of malarious disease of late years, and in the central portion of the Doab, or tract between the Jamna and the Granges, which is said to have got water-logged of late.* Even here the density is comparatively high, and we find about 375 persons to the mile in the former tract, and between 400 and 450 in the latter. The more thickly peopled part of the province lies first, along the line of rail, in the upper Doab, and still more remarkably in the South central group of districts, which includes all of Oudh lying to the west of the Ghagra river. In Table D., the two districts mentioned as con- taining, the most dense population are Benares, in the North-West Provinces, and Lucknow, in Oudh. Both these, however, are peculiar in being more or less appanages of great cities, and include, accordingly, a comparatively high proportion of urban and suburban inhabitants. We may take, therefore, as illustrations, more rural tracts, and in Balia, Azamgarh, and Jaunpur, we find three, in none of which is the density less than 800 per square mile. Bengal. Crossing into Bengal, we find the same high ratio. ' In dealing with this immense population, the historical divisions of Bengal, Bihar, and Orissa, are found unwieldy, save in the last instance. Bihar has accordingly been split into two, and Bengal sub-divided into three parts. But for old association's sake, it is worth while to give the population of so well-known tracts. In Bengal, then, the population is returned as 40,398,265. ^ Bihar returns 21,265,150, and Orissa 4,047,352. To these must be added the later acquisition of Chiitia Nagpur, the Feudatory States in which have been described above. The population of the British portion of this tract is given in the marginal table, since it is sufficiently well marked ofi" from the rest of the inhabitants of the province to warrant the distinction. In the first place, of the seven districts containing a population less dense than the general mean, four are in this division. It is fortunate, again, in lying almost entirely within the zone of light but certain rainfall. Lastly, as has been said in 'the Introduction, the hilly tract of which it is chiefly composed is the last refuge of the dark races that were expelled from the plains by the foreigner, or absorbed into his ranks, if they elected to stay there. Two of the three remaining districts of light population are of a similar description. One is in Orissa, and is the inheritance of the Khond tribe, of whom much has been written since the British occupation of that tract and the adjacent Madras districts. The other lies in the wild country separating Bengal from Burma and consists of a series of forest-clad hills and deep valleys, of which the latter only are under cultivation. The remaining district in this category is Darjiling, originally part of the Sikkim domain, now a flourishing tea-planting centre, with a favourite sanitarium, and a considerable trade with the tribes of the surrounding Himalaya to the north. But the arable area, except in the submontane tracts, is scanty, and the population sparse, as it is in the district touching Darjiling, in the plains. But the rest of the division specified in the marginal table as Northern Bengal, lies in a partially alluvial tract, with a density of more than 500 per mile. Here, however, begins the influence of the great rivers, which by the gradual shifting of their course have wrought most serious damage to the country, both in respect to its fertility and to the health of its population, through the obstruction of the natural drainage. The next division, Eastern Bengal, is still more subject to the river influences and Tract. Dis- tricts. Area. Density. Eabfall. r Northern 8 19,253 459 85''73 < Western 11 26,491 562 58-11 |_ Eastern 10 32,165 531 82-59 f Nortli Bihiir - |_ Sou til „ 5 17,515 667 51-82 5 19,300 520 45-14 Chiitia Nagpur 5 26,966 134 46-84 Orissa 4 9,853 411 58-86 * The remarks on these two tracts must he taken as suhject to correction, since they are based on th*. provisional report of the Census Superintendent, whose final wo)-k has not yet been submitted. 39 throughout the littoral tracts, communication is chiefly by boat, all the year round. Thi^ continuous , supply of fresh deposit, with the albundanoe of wholesome water, make the tract one of health and plenty. It not only feeds its own population of over 500 to the mile, and annually augments it, but it exports also large quantities of rice both to the less fortunate parts of Bengal, and to Madras and elsewhere in India and "abroad. It has been epigrammatically stated that in Bengal the population is in exact proportion to the productive power of the soil, measured m rice, and there seems good ground for the saying. As regards the density of this tract, however, we must make allowances for the inclusion of the hill tract of Chittagong above mentioned. The deltaic waste-lands of the Sundarbans, which fall within it are apparently excluded. Its limits, too, are being gradually restricted by the encroachment of the estuaries of the Ganges and Brahmaputra to the eastwards. In Western Bengal the metropolitan area has been included, with its population of nearly a million. There is not here and in Madras, as in the case of Bombay, that great distinction between the two classes of population which would affect the ratios of the whole surrounding tract. Ill the latter, the mainland is comparatively sterile and thinly peopled. It is hot so in the other two instances. The great distinction found in this division, and it is one that justifies still further partition, ethnologically, is that between the country on the east and that on the west of the estuary we know as the Hughli. Even in connection with the present subject, there is a noteworthy rise in density as we pass over to the eastern bank, and the western portion tends towards the characteristics of the Central Hills, both physically and in point of racial admixture. Orissa, apart from the hill tracts, consists of two belts ; one salt and less fertile, the other, one of superb rice-growing soil. The density varies, accordingly, a good deal in small patches throughout the plains, but in the main supports the same specific oopulation as the Noi-th-West Provinces, without Oudh. Bihar has here been divided into the two main sections of north and south, but there appears to be fair ground for further sub-division, as each of the above contains a zone of extremely dense population, lying along the Granges, with a fringe of comparatively lightly-peopled country, situated under the Himalaya on the north side of the river, and along the foot of the Chlltia Nagpur plateau and in the Rajmahdl hills with their surroundings, on the south. North Bihar contains the highest densities in the whole province, and in some places these have crept up almost to the sub-Himalayan forest. In the southern division, except for the influence of the great towns, there is a lower specific population, and the belt of almost thinly-peopled country is far wider, and is continuous throughout the tract. It is here that the population has to some degree outgrown its means of livelihood, and the seasonal migration to Calcutta, with the importation of large supplies of rice from Eastern Bengal, shows this. On the other hand, north of the river, in all places where there are no special disturbing causes, such as river changes or malarious outbreaks, the greater density seems to be supported without difliculty, and, as will be shown in the next chapter, it has not reached its limit. It is as well to mention this fact here, as otherwise the density might seem incompatible with the prosperity, or even the maintenance, of the population at its present level. The whole cousatry has now been reviewed with reference to the density of its Pressure of population, that is, merely the number of people to the square mile. Behind this, population, however, lies the more important question of the pressure of this population on the land. In order to appreciate this it will be necessary to travel considerably beyond the mere census record, so that the subject can only be treated here in the barest outline. There are instances in which the temporary migration of a portion of the inhabitants, together with the continuous inward course of the traffic in food supplies, are facts which need no corroboration. But in most of the cases which call for examination, the existence or the absence of congestion of the population depends upon many less patent factors. For example, there is the ratio of the inhabitants to the arable area, which in its turn is qualified by the relative facilities of communication with local markets for agricultural produce, the capabilities of the soil, the character of the cultivating classes, and the relative variety of the field of employment. As to the first and main point, even in British India, to which the following remarks are confined, the record of arable land, especially if that in occupation be distinguished from the remainder still available for fresh tillage, is by no means complete, owing to differences in the tenures prevalent in the various provinces, and in the system on which the unoccupied arable area is administered by the State. The figures of popu- lation can thus only be applied to the statistics of agricultural areas by experts in matters of , land-revenue, and on this consideration the census authorities deprecated E 4 40 N.-W. Pro- vinces and Oudh. the inclusion in their returns of any statements in which the two were combined. The omission has been supplied, as was suggested, by special investigations through ex- perienced local officers, from the results of which much more useful information has been obtained than can be incorporated in the present work. The general statement made at the beginning of this chapter, regarding the divergence of pressure from density, is borne out in several remarkable instances. The most striking, no doubt, Bengal. ig that of Bengal, where we find a very dense population supported in the eastern division with ease, for the soil is alluvial, continually renewed, well watered with a rainfall distributed over nearly three- fourths of the year. Passing over the tracts accidentally afflicted of late years by epidemic, we reach, in Bih^r, one tract with a density of "930, of almost proverbial prosperity, due not only to its natural fertility but also to the admirable system of roads by which it is traversed, to the capabilities of the tract for specially remunerative crops, such as indigo and the poppy, and last, but most of all, to the very high class of its agriculture. Adjacent to it are two districts with densities of 902 and 840 respectively, in which there is, no doubt, con- gestion, due to the subdivision of the arable area beyond the capacity of the cultivator to get a living out of it. It is said, however, that the competition for field labour is keen, so that one of this class, whether he is landless or not, is better off than a man who confines his attention to a patch of less than four acres for the support of five or even six persons. There is one district here, however, Ohampdran,- which is reported to be still under-peopled, though it returns 526 persons per mile. To the south of the Ganges we come to another congested tract, distributed, however, over parts of three districts, where many of the holdings are of two-and-a-half acres only. Here migration for a part of the year is habitual to a considerable part of the population, who find a living in the Calcutta Aills and the rice fields of the east, and even return with savings from their annual trip. In the North-West Provinces and Oudh it seems difficult to point to any special local congestion ; but if there be any, it is most probably amongst the higher social grades of small cultivators in the central parts of Oudh. The case of the less fertile districts in the central Doab is hardly one of congestion so much as of bad luck, susceptible of remedy. But the main point in connection with these provinces is the possibility of general congestion, since the area available is comparatively small, the population is large and very evenly distributed in the plains, so that it is necessary for the total annual surplus of food stock to rise proportionally with the increase of the people, as importation of food to an agricultural community of this size is almost out of the question. The contingency, however, has not yet arisen. Paiijah. In the Panjab there seems to be no congestion, except in a few parts of the sub- montane tract. Four districts show a remarkable density of population on the arable area, Hoshiarpur, Jalandhar, Sialkot, and Amritsar ; but all are within reach of the new field for cultivation thrown open by fresh irrigation projects. In two other districts the ratio to the arable area is very high, but it appears that these tracts, Hazara and Kangra, are largely pastoral, for which pursuit the vast expanse of hill- slope affords ample room. Sindh and Sindh, with its System of cultivation by irrigation, is not likely to feel pressure Bombfiy. of population so long as the Indus continues to run from lake and glacier to the sea. In the Bombay Presidency there are two tracts that may be said to feel the pressure of their population, and neither is densely peopled. In the Eatnagiri district, below the Ghats, there is congestion of the sort mentioned in connection with some tracts of South Bihar. The area suited for cultivation is small and nearly all taken up ; millets are imported from the adjacent Deccan tracts along the top of the Ghdts and a considerable number of the lower agriculturalists annually migrate to and from Bombay, where there is always work for them in the cotton mills and docks. The density, however, is but 282 per mile. In the case of the Deccan, the primary cause of the pressure is the uncertainty of the rainfall. There is also the fact that the richer classes of soil are comparatively small in area, and are occupied, for the most part, by the better class of cultivators. There is thus a goodly proportion of the population that lives off the lighter class of land, and is accordingly very sensitive to the variations of the season. The condition of the people of the northern part of the Deccan was made the subject of a special inquiry last year (1892), and confirms the above view regarding the influence of the season on the finances of the cultivator. The density of the tract in question lies between 245 per mile in the most westerly district, and 140 in the north and cast. In Gujar^th, the more densely populated division of this Presidency, the pressure is said to be scarcely felt. 41 Madras, too, has its most unfortunate tracts on the .Deccan plateau, and the Madras, remarks about the Bombay portion of this tract apply to Karnul, Anantapur, Belari, and Kadapa, where, in Karnul, the density of 109 is too great, as 134 is in Anantapur. On the other hand, Tanjore, with 600, does not yet show signs of pressure, owing to the extensive and fertile rice tracts under the Kavari irrigation system, and the facility of making short and lucrative trips to Ceylon for labour during the season when- the fields do not require the raiats' attention at home. ^ The Central Provinces have generally been regarded as a field for immigration Central from the more densely peopled tracts of Bengal and the North-West Provinces. But Provinces, if we deduct from the area surveyed the 20,000 square miles that are included in State forests, the density rises to very nearly that of Bombay. Taking, as has been done in the recent inquiry into the condition of the people, only the area which is under cultivation, it will be found that in the so-called " black-soil " tracts, where, as a rule, only winter cropping is practicable, there is but little room for expansion, and this tract ojitends over much of the north and west of the province. In the rice districts, as in Bengal, the densities are far higher and the pressure less. The general statistics of Berar seem to speak for themselves. There is com- Berar. paratively little arable land outside forests that is not taken into occupation, but less than 40 per cent, of the area under crops is devoted to the growth of the staple food of the province. The out-turn, even in a year of short rainfall, is enough to supply the whole population, and in an ordinary year over 40 per cent, of the crop is available for export, in addition to the produce specially grown for the foreign market, such as cotton, wheat, and oil-seeds. The two small provinces of Ajmer and Coorg need few remarks. In the former Ajmer and the railway has largely increased the town population and the means of feeding it. In Coorg. the rural tracts the precarious rainfall and the small area of arable land not taken up seem to indicate a certain amount of pressure, but on the other hand the. reports on the appearance and the fecundity of the cultivating classes prove that the limit has not nearly been attained. In Coorg, too, in spite of the area under non-food crops, the native population is easily, supported on the plots of hill land that alone are available in much of the province, but during a good part of the year the immigrants from Mysore and the coast are numerous, and the, local supplies turn out insufficient. Pressure, however, cannot be said to exist at present. In Assam proper there is rather a want than an excess of population on the land, Assam and and even in Sylhet, with its 398 persons to the mile, there is no pressure, and food Burma, leaves the district every , year. In Burma, as in Assam, there are thousands of acres of rich soil awaiting cultivation. The annual exports from Lower Burma include as much rice as is used by the whole population, including the supply occasionally required for the central districts of the Upper Division, when the rainfall is short there, which is a not uncommon occurrence. There is not, however, any sign that the agricultural population is pressing at all on the arable area. From the above sketch it may be gathered that the agricultural class, that is, the General bulk of the population of British India, is not pressing too closely on its means of sub- conclusions, sistence, except in a few special localities. In a part of Bihar, and in a small tract on the Bombay coast, the local produce is insuflScient for the inhabitants every year, and migration is adopted as a mode of relief. In parts of the Deccan the uncertainty of the season makes a pressure felt in circumstances where otherwise the land could bear the burden with ease. In the centre of the Gangetic valley there is a zone in which the margin of subsistence has been nearly reached, owing to not simply the weight of numbers thrown on to the land, but in part also to the liability of the tract to failure of rain. With these exceptions, the local resources suffice, and in several parts of India are year by year in excess of local needs, so that a large stock is always available for transport to places where it is wanted. Such is the problem in the present day, so far as the great mass of the people is concerned, and its further development will be discussed, so far as the census returns afford grounds for conjecture, in the next chapter. It may be noted that in the foregoing review the pressure of population has been Pressure in regarded from the standpoint of the agriculturist, and looking at the heavy prepon- ^^^"°{[^^JJl derance of this class with that immediately allied to it, the village menials, the attitude ofcupations. is justifiable. -The two together comprise little less than three-fourths of the whole community. But setting aside the trading class, which rises and falls in great measure with the condition of the masses, it is necessary to add a few words as to I 78388. ■•^ 42 those of the professional classes and such artisans as are not semi-agriculturists under the village organisation so commoji throughout India, There is no doubt that the general rise in prices of food, which is to the gain of the cultivator, has hit these two classes rather hard, and what with the supply of the former being in excess of the present demand for professional services, and the concentration of much of the textile and metal industry into the large factories in the towns, there is the possibility and likelihood of both professional and artisan falling back on the land for support, not so much as agriculturists as exploiters of such estates as they may be able to ^ obtain. In connection with what has been said above on the pressure on the land in some parts of the country, this prospect is not to be regarded as of no significance, though it may happily stillbe very remote. Definition of urban popu- lation. Town and Country. The stage of social aggregation at wrhich a community ceases to be rural, and passes into the category of urban, can nowhere be sharply defined. 'In no two countries is the line drawn on the same principle, so that comparison of the respective classes has to be confined to the places the population of which is assumed to be universally a guarantee of their urban character, and the respective proportions of urban and rural, as a whole, have been voted to be beyond the scope of international statistics. As regards the smaller aggregates, the population standard is liable to be as delusive as the constitutional test, for their size depends very much on the density of the country and the physical resources which determine the bent of the occupation of its inhabitants. The title of town, again, is conventional in most countries, and applied in consideration of varying constitutional distinctions, such as those of city and borough, in England. ' In India the difiiculty of classifying these small places is peculiarly felt in the present day, when the rapid extension of railways and other means of communication brings with it a very considerable amount of shifting of the trading and mechanical communities from place to place. It becomes necessary, therefore, to adopt three general tests with reference to the smaller units of population. First, that of constitution, that is, has the place been established as a municipality, or brought under some similar regulation for police and sanitary purposes ? Secondly, if neither of these methods of local' government has been applied, is the proportion of the trading and industrial population to the total equal to, or greater than, that of the agricultural? In the latter case the general numerical standard of 5,000 inhabitants was prescribed, as experience shows that taking the whole country together, this represents about the limit of urban preponderance. If .the numerical standard alone had been adopted, many of the smaller municipalities and other towns in Northern India and Mysore would have been dropped out of the category of urban, and, on the other hand, a considerable number of the large aggregates of homesteads on the Malabar coast, which are merely revenue units of a purely agrestic nature, would have come on the list. The system of classification of these comparatively small places obviously leaves the door open to a good deal of variety of interpretation of the term "urban" as applicable to the population thereof, but, on the whole, the results show that local experience, at all events, in British territory, was usually right as to the preponderance of non-agricultural occupations in the places determined to be towns, and the return shows that no less than 505, or nearly 25 per cent., of the total number of towns fall below the prescribed standard in point of population. The adscription of so large a proportion of the people of India to the soil, a point that has already been brought prominently forward in the foregoing part of this chanter, is in itself a sufficient explanation of the remarkable sparsity of the urban population. The village system of organisation will be treated of below, and where the rustic finds a scheme of existence so thoroughly adapted to his wants and aspirations, where rank is hereditary, and custom provides a regulation for every contingency, the oiTers no attractions, save to the AduUamite and Nonconformist, and the villager sticks to his surroundings, Porumque vitat, et superba civium Potentiorum limina. Distribution We find, then, out of the 717,549 places returned at the census, only 2,035 classed of urban us towns, and the rest under the head of villages. The urban population is in the population, proportion of 9-48 per cent, to 90-52 of rural. In British territory, the proportion falls to 9-22, and in Feudatory States it rises to 10-38 per cent. The probable explanation of the difference will be found in the following statement, F, where a city average 43 summary is given of the main features of the urban population in the different parts of India : — . ' /^ F.— -DlSTRIBUTrON^ OF THE ITeBAN POPULATION. Urban Population. Percentage of Urban Population in Towns of Mean Province or State. Total, Pfercentage on Total Population. 50,000 and over. 20,000. 10,000. 5,000. Under 5,000. Proximity in Miles of Towns of 20,000 and over.* Madras 3,406,105 9-56 29 24 22 23 2 1 68 ■ Bombay Sindh 3,116,304 342,295 19-49 11-92 ■ 43 13 22 17 5 1 " L 125 Bengal - 3,443,876 4 -.82 44 24 22 9 1 78 "N.-W. Provinces 4;352,573 12-70 38 11 17 21 13 56 . Oudh 961,755 7-60 36 7 22 25 10 75 Panjab - - 2,413,704 H-56 41 15 15 22 7 81 Central Provinces 739,952 6-85 27 20 23 .30 — 149 r Upper Burma 371,404 ]'2-60 51 — . 17 27 5 -1 [Lower Burma 575,245 12-35 • 41 21 20 13 5 168 Assam 102,074 1-86 — — 34 43 33 , — Berar 360,711 12-45 — 25 28 47 — 83 Ajmer /, 118,631 21-87 58 36 — 6 — 32 Coorg 15,511 8-96 — — — 45 55 — Haidrabad 1,090,129 9-45 38 10 17 35 — 138 Baroda - - , 483,515 20-02 24 11 30 31 4 56 Mysore 626,558 12-67 41 — 9 18 32 127 Kashmer 197,743 7-77 60 18 5 17 — 216 Eajputana 1,530,087 12-73 26 14 25 35 — 117 Central India 964,538 9-34 28 25 17 30 — 87 Bombay States 1,177,422 14-61 5 29 26 31 9 * Madras States - - . 175,125 4-73 — 29 42 29 — * Central Province States 38,656 1-79 — — — 100 — * Bengal States 16,542 0-50 — — 69 31 — * N.-W. Province States 103,188 13-02 74 — — 16 10 * Panjab States 456,544 10-71 12 9 24 39 16 * Detached Settlements 71,349 — — — — — — — Total India - 27,251,176 9-48 35 16 20 23 6 90 Number of Towns 2,0 3 5 — 78 149 407 896 5oS 3o * The smaller groups of States are in this column included in the Government with which they are connected. One striking comparison suggests itself which is not brought out in these figures. Proximity of It is that whereas in England 53 per cent, of the whole population is found in the 182 *owns. towns of 20,000 and upwards, in India, though there are 227 such towns, only 4-84 per cent, of the population reside in them. The limit of 20,000 has been selected as being about the smallest that may be presumed to have any considerable influence on the rural tract of which it is the centre. The last column of the statement, accordingly, shows the average distance between one of these towns and the next, on the assumption, of course, of equal distribution, as is the rule in such computations. It thus appears that, taking India as a whole, we shall find a town of not less than 20,000 inhabitants every 90 miles. If we exclude the larger States, the distance is somewhat less, owing to the removal from the calculation of the large areas of Mysore aigd Kashmir, which boast only two towns each of this size. Burma and the Central- Provinces then show the worst provision in this respect, but it must be remembered that in the latter the F 2 44 Proportion of urban population. Large towns. outlying States have been taken into consideration, since most of them use the towns in British territory. In the case of Bombay, the difference caused by the addition of the States is hardly felt, as the latter contain a fair number of towns of the requisite population, whereas there are none in the small States of the Central Provinces. Baroda and the North-West Provinces are the best off for large towns, since Ajm^r may be left out of the account, as its circumstances are peculiar. Madras, too, comes high up in the scale. Bombay, Oudh, and Bengal, are all much on an equality. The next point is the proportion of urban population to the total. Again omitting Ajmer with its three railway markets, the highest percentage is that of Baroda, which, in addition to the capital of the State, has a fair number of small towns. Bombay is the next in order, and, with its States, is considerably above the rest. Sindh, which has been taken with it in the table, shows only 12 per cent., in spite of its four large towns. The proportion in the Panjab, too, is much the - same. In Bengal the urban population, which is of itself considerable, is utterly swamped when taken alongside of the enormous mass of rustics in this province. The pro- portion in Upper Burma seems high, but half of the number returned is found in Mandalay, and there is a long drop from the capital to the next'town. It would be much the same in Lower Burma, but for the influence of Maulmain, the younger sister of Eangoon. Kashmir, too, shows in the return the predominating position of Shrinagar in its urban element. On the other hand, Madras, without, any specially large town besides the capital, has a wide distribution of middle-sized places, including several flourishing seaports. The North- West Provinces do not appear to due advan- tage in the return, since the proportion of* their numerous large towns is outweighed by that of the still more numerous small ones, which have been of late years brought under the local regulations regarding police and sanitary provisions "without being constituted municipalities. In Mysore, too, it seems that something of this sort has been done. In several of the States we find one large town with a few much smaller ones. This is especially the case in Rampur, of the North-west group, in a few of the Panjab States and in most of those in RAjputana, and finally in Haidrabad. The urban element, in conclusion, is at its lowest ebb in Assam and the small Hill States of the central belt. Another feature in the return is the comparative paucity of towns of large size. Out of the total tale of 2,035, there are 1,401 which do not contain as many as 10,000 inhabitants. Between this and 20,000 there are 407, whilst above this limit there are, as we have seen, 227. The towns of 100,000 and over number only 28, or 30, including two of the suburbs of Calcutta to make the total completely in accord with the general tables. The following statement shows these places, in order of magnitude : — 1. Bombay 2. Calcutta, &c. 3. Madras 4. Haidrabad 5. Lucknow 6. Benares 7. Delhi 8. Mandalay 9. Cawnpore 10. Bangalore 11. Rangoon 12. Lahore 13. Allahabad 14. Agra 821,764 741,144 452,518 415,039 273,028 219,467 192,579 188,815 188,712 180,366 180,324 176,854 175,246 168,662 15. Patna 16. Poona 17. Jaipiir 18. Ahmedabad 19. Amritsar 20. Bar^li 21. Meerut 22. Shrinagar 23. Nagpiir 24. Howrah 25. Baroda 26. Siirat 27. Kardchi 28. Gwalior 165,192 161,390 158,905 148,412 136,766 121,039 119,390 118,960 117,014 116,606 116,420 109,229 105,199 104,083 The great gap between the three Presidency towns and the rest is .filled, it appears, only by Haidrabad, in which is included the large suburban cantonment of Sikandrabdd. As regards the claim to be the first city in India in population, it must be considered that Bombay is under one municipal body, and contained within well- defined boundaries, and has no suburbs on the island itself, which means none within about 12 or 14 miles. On the other hand, the extension of Calcutta is like that of London, laterally ; so that whilst the nucleus of the metropolis is much as it was 10 years ago, the outskirts are growing into suburbs in all directions. There is again, the large town of Howrah, across the Hughli, which, though in another district and under separate municipal government, is as much a part of the city as Southwark IS of London. Taking all these straggling units together, the metropolitan population rises to 961,670, under no less than six municipal bodies, with three different 45 Presidents. Irrespective, therefore, of the outskirts, Calcutta must be content witli the second place, both in population and iu homogeneity of administration. Madras covers a large area, and comprises within its limits several collections of houses that are almost separate units ; but the whole is under one municipality, and there is no question of suburban aspirations. The floating population, too, is small compared to that bf the other two cities, since there is little of the manufacturing industry on a large scale that attracts so many of the labouring class to Bombay or the banks of the Hughli. The shipping trade, too, in Madras is by no means large. Haidrabad includes, besides Sikandrabad, a considerable suburban population, but it is in itself a very large city of quite the native Indian type. Lucknow, like the rest of these large towns, includes a considerable cantonment population, and in several other respects resembles Haidrabad, in spite of a generation of British occupation. ' The population of Benares, its neighbour, fluctuates considerably from month to month, as a large proportion of it consists of pilgrims. It is not necessary to go through the rest of the roll in detail. Bangalore, Agra, Poena, Bareli, and Meerut, are largely maintained directly or indirectly, through their position as great military dep6ts, or as the head-quarters of civil administration of a considerable tract. In some degree this may be said of Lahore and Allahabad. Mandalay, Jaipur, Gwalior, Shrinagiar, and Baroda, owe their position to the fact of their having been for so long the capitals of wealthy States. Nagpur is fortunate in having no rival near it, and in being surrounded by a fertile country, the produce of which finds its chief outlet through this town. Karachi and Eangoon are practically the creations of British trade. Cawnpore and Ahmedabad may be bracketted together as combining to a great extent commerce, of the more extended kind, with manufacture on the modern system. Delhi is the admitted metropolis of the north-Vestem and larger portion of the Gangetic basin, as Lucknow is of the north-eastern. The former has, moreover, entered into the movement of modern industry, and is not entirely given over to the system .on which it grew up under former dynasties. "We have, finally, Patna and SArat, both formerly provincial capitals of note, but failing- to keep up the population of 10 years ago. Their condition will fitly serve as an introduction for a few words on the diffierence Townward between the nature of the tendency to town-formation that we find in operation at tendencies, the present day, and that in pursuance of which most of the older cities of India were developed. The extent of the tendencies and the degree in which they are now in action are matters which appfertai^ to the dynamics of society, which will be treated of, so far as they fall within the scope of the census, in the next chapter. But as ofiering explanations of the existing condition of things they are relevant to our present subject. In the minds of the great majority of the masses, city life and its attractions are no more than a " nebulous hypothesis," and the town might just as well not be in existence. To the upper classes it is distasteful, saving to a native Chief in his own capital, of which more anon. The local magnate in his own domain is a Triton amongst minnows, by hereditary right, but in the city the equality of all before the law, a feature in British administration which he despises, is at its full height, and liberty lords it over birthright. Thus the field is left open to the trader, the pro- fessional, and " business man '"' generally, and under the influence of railways and foreign commerce his horn has been greatly exalted, to the prejudice of the others into whose presence, a few generations back, he could not have hoped to be admitted. There is also to \)e considered the stimulating presence of the foreign element found in the centres of trade, which is thrown into the scale in favour of the middle classes. On .the whole, then, the main factoi-s in the development of the cities of the present day, stch as the seaports. Presidency towns, and the few trading and manufacturing centres in the interior, have been foreign capital administered by foreigners, and the scope given to the talents of the native trading classes. To a minor degree, too, the artisan and professional has contributed to the growth of such towns. Let us now turn to the conditions of town life in India before British influence Court began to make itself felt, or to where, in later days, it is confined to merely the protective stimulus. functions of a paramount power. In former times, whatever the theory, in practice the State existed for the maintenance of the Chief, and his duties began at the outer edge of the frontier. The public revenue was sucked into the treasury, and the expenditure was limited to the army and the personal tastes of the chief, or the embellishment of his palace and capital. It has been said by Jacquemont, that in India the authority of the Chief " decroit au moins comme le cube de la distance du lieu oii il se trouve," so that what with the quasi-feudar relations of the clansmen amongst ■ the Hindus, and the system of payment by "jagir,"or the assignment of tracts to skin, by the Moghals, tjiere was little temptation to amass wealth that could P 3 46 not be protected, except under the immediate wing of the chief's courtiers. This state of affairs was fatal, of course, to any growth' of the smaller towns,* but excessively stimulating to the congregation in the capital of such professions and industries as ministered to the requirements of the court and army. The sole chance for an aspiring author or artisan was to attract the attention of the Chief or_ Emperor, as the case might be, or of the reigning favourite. To do this, he was obliged to go to the court, and, if lucky, to remain there till he had buried an ample fortune in some wall or garden. In fact, as Bernier says, " From the nature of the government of this '" country, a capital city, such as Delhi or Agra, derives its chief support from the " presence of the army, and the population is reduced to the necessity of following " the Moghal whenever he undertakes a journey of long , continuance. Those cities " resemble anything rather than Paris ; they might more fitly be compared to a camp. " The king's pay is the only means of sustenance." "We thus see how the dialect of Hindi spoken round the head- quarters of the Moghal came to get the name of Urdu, the camp language. It is not probable that in the older States, especially under the Rajput clans, the concentration of trade was as great as is pictured by the genial physician, as the peity sub-chiefs each made a point of establishing a little court in imitation of their superior, but outside these cities of refuge, the trader had no pro- tection and but little business. Then, again, looking at the nature of the demand in these towns, it appears that the greater and more lucrative traflB.c was in arms, ornaments, fine fabrics, cloth of gold, illuminated manuscripts,' and other object^, of what economists call unproductive expenditure. So factitious a stimulus could not long survive the regime under which it flourished, nor was it likely to be continued under the plain utilitarianism of the British system of government. Nowadays, whatever may be said by the denizens of the Presidency towns regarding the annual " exodus " of the Government to hill stations, the support, such as it is, to commerce, which is thus withdrawn, is insignificant in the mass of new openings that have sprung up independent of it. In proof of this the general condition of the chief towns may be brouglit forward. The seaports and the manufacturing centres, and the great produce-markets in touch with the coast, all of recent birth or regeneration, are flourishing. Of the former capitals, apart from the presence of military forces, inferior in number and luxury, perhaps, to those of yore, but better paymasters, Delhi, which, as above mentioned, has entered into the race on the new conditions, is almost the only one that holds its head up. Yijianagar lies in a boulder-strewn waste. Bijapur, that once sucked in the wealth of the Deccan for miles round, is now but' a skeleton into whose dry bones life is just returning under the magic touch of the iron horse. Agra and Dacca just hold their own, and that is all. Patna fails to do this, and Murshidabad is little more than a large village, overshadowed by the palace of those who were once gracioijsly pleased to allow Calcutta to exist, but forbade it to grow. Lastly, look at Surat, the port of the Empire, the Gate of Mecca, which, with its foreign relations, was the only place that showed signs of taking rank with modern emporia. Its river has played it false, and it is now hopelessly stranded in the backwater of commerce. Of the latter-day towns that are still in motion, more will be said in the next chapter. The village A wider, and a more important subject is opened out in the consideration system. ©f the distribution of the rural population. There are few topics connected with India that more repay investigation than the different systems on which this population has formed itself into the microcosm which we know by the generic title of village, but it will be out of place here to enter into more than the outlines of it. In the first place, it should be explained that for the purposes of the census, the term village is used in most parts of India to denote the unit of administration of the land revenue, and this implies, as a rule, the existence of a congregation of the inhabitants on a fixed site, with occasionally some outlying hamlets. It is not so, however, in parts of Lower Bengal and of Assam, where the land settlement, not being subject to periodical revision, does not require the maintenance of minute administrative records, and the population scatters itself as it pleases within the limits of the village. It is not so, again, in most of the hill tracts, where the boundaries of the village, as a unit, may be surveyed or not, but the inhaMtants thereof are not to be bound down to a' fixed site, the spirits attached to which may at any time turn out suddenly malignant, leading to unexplained death or sickness in the resident families, or, still worse, to the loss of some of their fowls or cattle. In such localities, therefore, the tie to the site As this was passing through the press, Hofrath G. Biihler, C.I.E., to whom I referred this point commu- nicated a few instances, such as Pali, Nagara, and Siddhapur, in Western India, where a town flourished other^vise than under a local chief. There are doubtless others, but they seem to make the general rule stand out all the more clearly. — J. A. B. 47 is a very lax one, and it takes little to break it. Then, again, on the Malabar coast, the Yillage, such as it is on the plateau above the Grhats, is almost unknown, so the unit taken is the collection of separate homesteads, known as the Desham, which corresponds to the Deli in the desert tracts of Sindh. When, too, we come to the constitution of the village, we find as much difi'erence as there is in appearance between the walled and castellated mass of buildings that repels outsiders in the Panjab and Decoan, and the unprotected bamboo-covered group of cottages that rises out of the rice swamp of Lower Bengal. The title of village community, which may be used of all, must not be held to imply that all possess the special feature so thoroughly described by Sir H. Maine in connection with the corresponding communities in the west, namely, the joint property of the whole body in the lands of the unit. This exists, though often in some modified form, in the north of India, and is the pre- dominant form of village organisation in most parts of the PanjAb, Here the question of race comes in, as the communities from the north-west who have from time to time made their way into India across the mountains, have generally colonised the frontier Province in force, but have not strayed, save in small bodies, further into the plains. The great exception is that of the race known as the Aryas, who avoided the Paniab and fixed their abode lower down the Granges valley. Lastly, there is to be considered the race, or races, of darker colour and lower civilisation who were displaced by the immi- grants, whether Arya or other. Inquiry stops here. Amongst these, who are generally called the Dravidian race, a term which may be used here for want of a better, it appears that the joint proprietary village was scarcely, if at all, recognised, though there were leaders of the cotnmunity who held a predominant position in the village without interfering with individual property in the lands thereof, but who regulated the dealings of the community with the waste land and uncleared forest. It is possible that in most instances to this class was originally given a colonising or clearing lease of the village by the superior agent of the tribal Chief, who had the grant of a larger tract, which was subdivided as above. This is very different from the position of the joint owner of the northern frontier. Here, he is in most cases the sharer of all tlie land of the village, waste and occupied. In others, though the greater part of the land may have been partitioned ofi" into individual holdings, it is subject, it is said, to occasional re-division, so as to give each cultivator of the village-family the chance of a share of the choicest soils, or, as is more probably the origin of the custom, to prevent any single set of occupants from gaining a permanent supei"iority over the rest of the sharing clan. It appears, however, that in this part of India, either the Dravidian element was not in existence, or that the repeated waves of immigrants of various races obliterated it in pre-historic times. In the tract occupied by the Arya, on the other hand, which lay to the east of the Five rivers, there are, no doubt, traces of the imposition of hereditary proprietary right above, the right of individual occupants ; but in the present day, partly owing to more modern 'influences, the joint or landlord system of village property is universal throughout the North- West Provinces and Oudh, excepting, perhaps, in some of the hill tracts. Still further to the east, again, there is in Bengal, though not in Bihar, a village system created within the past century by the early British administrators, which, in the sense in which the term is used in connection with our present subject, is no system at all. The archaic form of holding -these estates was stamped out wherever the Moghal got a fair grip on the country, as was the case with the half of Bengal that was populated at the time it passed into the hands of the British. In this provinqe, accordingly, we do not find the more or less complete village organisa- tion that distinguishes this community in the rest of India, whether joint or individual. In the Central Provinces, too, the system of village tenure is in great measure the creation of British administration, but difiering from that in Bengal in the very important provisions made at the later settlement for the maintenance of the rights of the individual occupants against the encroachment of the newly-made proprietary title. Further south, the iron heel of the Maratha was exercised, like that of the Moghal in the north, in equalising subordinate rights, but the destruction of the former system was by no means so complete as under the more scientific procedure of the foreign absorbent of the revenue. Isolated instances of joint property are found in the west of India, one of which, in the north of Gujardth, resembles in almost every particular the joint system of North-west India, whilst another is singular in the fact that the sharers are not necessarily connected by caste or blood. In a third we find the revenue system of the Marathas perpetuated by the .creation of what has now been confirmed as a proprietary title, that is, of a kind of farih of a village, or collection of villages, with the preservation, as in the Central Provinces, of tenant rights to occupancy at settled rates, instead of at will. One of the last distinctions between F 4 48 the joint and the individual system of village tenure whicii it ie necessary to mention here, is the existence in the latter of the headman, as the leader of the community, and not, as in the former, as merely a State oflBcial for "the collection of revenue and the minor duties of administration. Under the joint system such an exaltation of the status of any one member of the proprietary body, would be inconsistent with the oligarchical rule implied by the tenure. In the southern system, on the other hand, the post belongs, if not hereditarily, at least by convention, to certain families, on the condition that the occupant of the post is in possession of a landed estate in the village. The village Such is the constitution of the village community, and the next point is its community, internal organisation. The Government is practically oligarchical, whatever it may be in name. If there be a headman, he acts as one of a body of " leading families." Where there is a coparcenary body, the village affairs are managed by a jury of the heads of the sharers. Megasthenes, on his travels, says that he found the institutions of the Panjab democratic, and perhaps they were _in those days, so far as the outside world is concerned. But the ambassador had been bred under democracy, and was predisposed by his training to.find traces of it everywhere. We only see what we have been taught to look for, and the first judges of the Calcutta Supreme Court found in the absence of clothes on the part of the boatmen who carried them on shore, a proof of the poverty of the country, which poverty it was their duty to eliminate, whilst the British soldier is apt to attribute the want of muscle amongst the Indian peasantry to the unsatisfactory distribution of butchers' meat. But in fact, to the average villager, the notion of equality is scarcely "thinkable," and to' those of higher rank it is abhorrent. If there chance to be a storm in the even tenor of country life^ directed against the ruling body of the community, it will generally be found that the motive was, as Sully put it, not I'envie d'attaquer, but I'impatience de souffrir — some violation of the immemorial custom of the village. In its economic aspect, the village aims at being self-sufficing, as may be assumed from its history. For not so very long ago, relatively to the life of a people, every village had to be in a state of continual preparation for defence against its neighbours. Community of interests was confined to the walls of the village, at least in the plains, and the ruling sentiment was towards peace internally, with distrust, if not actual hostility, in respect to all outside. Fo doubt, at a comparatively early stage in the development of the village in India, the local market was established for barter of surplus produce against articles of domestic use or convenience not to be made in the village itself, and these markets were regarded, according to Sir H. Maine, as the neutral ground of the inhabitants of the whole tract round. By-and-by, as matters settled down, room was made for the vendors of such articles within the village as residents, but not as members of the inner circle of the community, though they took rank above the groups of helots, who represent the displaced tribes found on the soil by the immigrant foreigners, and who were compelled to live in hamlets immediately outside the village site. Taking the " village community " of Sir H. Maine to refer to only the joint or northern Indian type of village, his work is still the standard authority on the subject, but, as shown by Mr. Baden- Powell, it -should not be taken as of universal application to the whole country. We find, then, that the community in question, in its normal stage, included onlv the differentiation of function necessary to supply the wants of an agricultural community, and, in the climatic conditions of India, these wants are few and simple. The spread of railways and other means of communication, and the general develop- ment of the trade of the country, have largely helped to add to the variety of the village population, but, as was remarked above, the new comers, in spite of their often greater wealth, occupy a social position considerably below that of the older members of the community. They make, accordingly, for the market town, so that out of the nine- tenths of the people of India, who, as shown above, dwell in villages, four-fifths at least belong to the classes that composed the original village community.* This is a matter that will have to be noticed in connection with the distribution of the population bv occupation, but it is so far relevant to the present subject that it shows how with the growth of the village m wealth and population, there is but little tendency towards as it were ; m the direction, that is, of needs of a sort-unknown to ^,^i^: * Where the tradition of hereditary title to the village lands is in its vigour the rank of a vator hailing from another village is far below that of the poorest of the " Old Standard " for instance, there is '^ •"--" "" •"'" • ' • extension laterally their forefathers. well-to-do culti- for instance, there is a proverb that "thePdllu (marriage processionrsoes"to "thT^nn"^' nf l^i^l''^*^"i^°' SltJl^^S^'^T^rr'^^'-)' P-* ^^^^ ^' ^'^ ^Tpri%eVeome;?4TliirSi: t:4i%nrr rolling in wealth).— J. A. B. ghi (is 49 The Statement Q. but the marginal note Group. Populatu n. No. of Villages. 20,000 and over 314,l"81 17 10,000- 1,455,214 109 5,000- 3,883,938 606 .3,000- 12,854,322, 3,469 1,000- 70,025,695 45,830 500- 67,475,109 97,846 200- 71,180,018 222,996 Under 200 32,625,858 343,052 Unclassed Total 20,478 — 259,834,813 713,925 country no doubt led to the north--west Panjab, again, the tended in some places to large, in the tract and the facilities for getting water establishment of small villages in favourable on the next page, includes both towns and villages, gives the latter separately, but without territorial subr division. It will be seen, however, that\ there are hardly any places of 10,000 in-| habitants and over which are not classed as [ urban in character. The few in the highest group are almost entirely confined to the coast States of Madras, where the unit has been taken to be, not the Desham but the collection of Desham, known as the Provarti, a system of grouping which renders the return valueless for comparison.- It must also be noted that the villages of the small State of Hill Tipperah, under Bengal, are omitted, for the reason already given. As regards the factors that determine the size of a village unit, a very few words will suffice. In former days the insecurity of a great part of the formation of comparatively large aggregates. In the tribal character of the occupation of the soil others to small, villages. Thirdly, the -fertility of where the rainfall is scanty, result in the circumstances, and of large elsewhere. Varying size of the village. In the hill tracts, such as those of Assam, the Central Provinces, Bengal, and north Madras the want of arable land limits the village population to a very low figure. In some tracts, again, the custom of splitting up the original community into detached hamlets is increasing, now that life and property are secure, especially- where the soil is so light that large holdings have to be taken up in order to provide for the family wants, and the cultivator would have to go long distances daily if he adhered to his ancestral homestead. Such offshoots are not, as a rule, counted as distinct villages, except in Bengal, east of Bihar, where the circumstances, as explained above, have been adverse to the formation of large communities. Then, again, in the Himalayan tracts of the Panjab, the village, in its ordinary sense, is not to be found, the hamlet taking its place. Thus the return shows that whereas in the British portion of this tract, the revenue unit has been used, in the Hill States the return recognises only the tika or hamlet. On the whole, the latter seems to be the preferable course to be taken in future, as the members of the tika are usually connected by some bond or other, but the village, consisting of a number of hamlets founded by entirely independent sets of cultivators, is a mere unit of record, the collective name of which is not always even known to many of the inhabitants, and certainly is not in colloquial use. At the other extreme of India, in Assam, there is the revenue unit in the Brahmaputra valley and in the upper portion of the Surma ; but elsewhere the people have squatted according to the con- venience of their field work, and the village, as a community, is unknown. In the hills, too, there is only one large tribe that dwells in fortified units of population, and the rest move about in the forest, as need or inclination dictates. In Kashmer, we find much the same tendency as in the adjacent Panjab hills to fix the abode close to the field, and as in such a hilly tract the area of land fit for cultivation is scanty and scattered, there is nothing in the shape of a large village except in the -valley. In parts of Sindh, again, where the distances to be traversed are very great, the village unit occasionally includes from 10 to 20 hamlets, distributed over many square miles of almost desert country. With the above explanations Statement Gr. can now be examined. The general mean population per village in India, as a whola, is 363. In British Provinces it rises to 374, and falls in the States to 330. But the exponential figures added to the mean of the chief Provinces prove that except in Oudh, the Central Provinces, and Berar, that mean is the result of very considerable district variation. In the case of the States, where such aid is wanting, one or two leading points of difference have been already explained. For instance, the high average of 2,727, in Madras, as compared with that of 667 for the British portion of the Presidency, is due to classification only. The same may be said of that in the Panjab States, which is 190, as compared with 532, for tbe divergence is greater in the Hill tracts than elsewhere, the reason being the different treatment of the tika or hamlet. I 78388. 0- Mean popu- lation of the villao-e. 50 Table G. — Showing the General Distribution of Towns and Villages. Average Per-centage of Population contained in Places of Per-centage of Places containing a Population of Province or State. Popula- tion per Unft. p,,_.l Proximity Po?Son °iJ"- %*« per Village, '^ffil™ 2''»- ' 200 500 1,000 5,000 . 10,00c 20,000 TT„j„„ . and Un.der ^oo. over. 2""- 500. 1,000 5,000 10,000 and over. MadrSvS 624 5ir^ 1 1-69 6-3 12-9 20-9 47-7 5-8 2-3 5-1 40-2 24-3 18-3 16-4 0-6 0-2 ■ Bombay 845 30-S1 605 2-05 3-8 15'9 22-0 37-8 5-4 4-7 10-4 24-6 35-4 23-3 15-7 0-6 0-4 Lsindh 766 ."■83 644 3-75 3'0 14-1 25'3 38-4 8-5 2-1 8-6 22-1 31*2 27-5 18-0 0-9 0-3 Bengal 314 4I'I3 299 0-88 16-7 30-4 24-0 23-4 1-1 1-1 3-3 53-6 30-8 10-9 4-6 0-1 - { N.-W. Provinces - 418 29-95 367 1-09 lO'O 25-1 25-3 28-2 3-1 2-1 6-2 44-6 32-4 16-6 7-1 0-2 0-1 loudh 519 ^P' 0-34 6-1 24-1 31-3 31-5 2-1 1-6 3-3 28-S 37-6 23-3 10-3 0-2 0-1 Panjab 599 iP' 1-92 5-5 19-1 23-9 38-5 4-4 2-2 6-4 31-3 34-5 20-5 13-1 0-4 0-2 Central Provinces 314 293 1-*1 16-1 35-2 26-4 16-1 2-4 1-6 3-2 150-0 34-8 11-9 3-2 0-1 ("Upper Burma 268 29-36 235 2-96 22-3 36-3 19-0 10-5 3-4 2-1 6-4 58-8 31-4 7-8 / 1-9 0-1 — (.Lower Burma 262 2J-82 230 2'39 22-9 39-9 17'7 7-9 1-5 2-5 7-6 57-4 33-9 7-2 1-3 0-1 O'l Assam 319 60*70 313 1-82 12-9 45-1 23-0 17-6 0-9 0-5 — 40-2 45-2 10-9 3-7 _ Berar ''- 496 I3'2I 439 1-87 7-6 23'0 25-8 31-1 5-8 3-5 3-2 37-4 34-7 18-4 8-9 0-4 0-2 Ajm^r 728 572 2-05 4-9 13-8 18-6 35-1 7-0 - 20-6 33-3 32-2 19-5 13-8 0-8 0-4 Coorg 348 320 1-92 U-2 32-9 3S'7 18-1 4-1 ~ - 45-5 32-6 17-5 4-2 0-2 - Total, Provinces 411 374 1-44 11-0 34-8 23-8 30-3 3-2 1-9 50 46-2 32-0 14-2 7-3 0-2 0-1 HaidrabM 573 522 2-18 5-3 20-8 29-1 35-4 3-3 1-6 4-5 27-8 36-5 23-9 12-4 0-3 0-1 Baroda 793 643 1-78 3-6 12-5 21-9 41-9 7-1 5-9 7-1 26-3 29-6 24-5 18-4 0-8 0-4 Mysore 293 257 1-37 19-3 34-3 21-0 17-6 2-5 1-2 5-1 65-6 32-1 9-3 2-9 0-1 _ Kashm6r 306 282 3-36 17-3 31-6 23-6 19-8 1-9 - 6-0 54-8 30-7 10-5 1 3-9 0-1 — EajputS,na 395 346 2-24 11-8 24-4 22-6 28-6 4-4 3-2 5-0 SO-0 30-2 13-0 6-4 0-3 0-1 Oentral India 317 289 1-68 16-6 31-2 21-5 22-5 2-7 1-6 4-9 54-0 31-6 10-1 4-1 0-1 0-1 Bombay States 522 449 2-29 7-5 20-2 25-2 33-0 4-9 4-1 6-1 38-5 32-1 18-9 9-9 0-4 0-2 Madras States 2^31 2,727 - 0-4 2-5 5-8 22-0 25-1 34-9 9-3 8-8 20-6 22-7 30-0 9-4 8-5 Central Province States - 207 204 1-80 27-6 41-0 20-2 9-4 1-8 - - 65-4 27-0 6-3 1-2 0-1 _ Bengal States 175 174 1-49 32-0 35-6 18-7 12-2 1-1 0-4 - 71-6 21-6 5-3 1-5 — __ N.-AV. Province States 442 298 1-60 12-0 45-1 19-2 11-9 2-1 - 9-7 38-2 48-1 9-8 2-8 0-1 Panjib States 212 190 1-48 18-9 20-7 21-2 30-0 4-3 2-6 2-3 76-0 13-6 6-4 3-8 0-1 0-1 Total, States ,- i 367 330 1-83 12-7 26-7 22-4 26-6 4-7 4-1 4-8 53-1 28-7 12-1 5-8 0-2 1 0-1 INDIA 400 363 1-59 11-4! J4-8' J3-6' 29-4 3-5 1 2-4 5-0 47-9 311] L3-7 70 0-2 0-i 1 Villages only - 363 1-59 I2"6 27-4 26-0 31-8 I'S 0-6 0-1 48-1 31-2 I3"7 6-9 o-i - 61 Panjab. British. States. Hill Tracts Submontane Tracts Salt Range Tracts "Western Plains Eastern Plains - - - 828 516 547 '■ 440 666 60 474 616 419 But in other parts of Panjdb. the Panjab, also, there is little agreement between the two, except in the submon- tane tract, in which the village is less populous than where the arable area is less productive. Baroda, again, shows the high average of 643, the same as that of Sindh, but the Baroda. cause is less artificial, as the size of the village seems determined chiefly by the circumstances of the northern division of the State, the soil of which is light, and the area under each unit relatively large. In the most fertile division, omitting the capital city, the mean is a little below that of the adjoining British districts, where the forest belt along the eastern edge, is narrower, and has less lowering effect on the average. Baroda. state. Southern Northern Western 535 853 485 In the case of the two coincidence of the averages North- West Provinces and Oudh. British. states. Hill Tracts Submontane Tracts - Upper Doab Central Doab North Central South Central Southern 106 375 565 501 416 425 363 301 297 States under the North-West Provinces, we find a curious n.-W. Pro- , in spite of the very different density of the two. As, vinces and however, the British hill tract shows only 106 Ou^h. per village, and the Hill State gives over 300, it is clear that the method of classification differs, as in the corresponding tract in the Panjab, only here, the tika is apparently recognised in British territory, and the village in the State. In the rest of the province the mean of different tracts is less uniform than usual. It is low in the southern hills, rises a little in the densely peopled south -central division, and very considerably in the Upper Doab, where the population, though still thickly spread, has to undertake the cultivation of a wider area. The Bengal States are chiefly in the hill tracts above mentioned, where land is Bengal, scarce, and the population is obliged to move about more frequently than in the plains, in search of fresh fields. On the eastern frontier, too, shifting cultivation is the rule, and when a forest tract has been wastefuUy burnt down for ash manure, by which the harvest of a single year is raised, the whole settlement migrates elsewhere for perhaps 10 or 12 years, by which time the low scrub has probably grown up again sufficiently for a second course of destruction. In the British portion of the hills and on the plateau of Chiitia Nagpur, the average size of the village is a trifle less than in the States. The highest average of the Province is found in North Bihar, where the village organisation is on the same footing as that of its neighbours in Oudh and the North- West Provinces. The high average of Eastern Bengal is in part attributable, probably, to the paucity of elevated building sites available for that amphibious population. In the only State found in the plains, Koch Bihar, the average is higher than even in Bihar itself, a result due, perhaps, to classification of hamlets in the sub-Himalayan portion of the State, but no explanation is given in the returns. The case of Assam requires little comment. The two valleys, in spite of their Assam. different densities, show nearly tbe same average population per village, but in neither is the village a compact unit, as in the rest of India. In the hill tracts, the size of the village is determined, as in the hills of Bengal, by the necessities of agriculture, and is very nearly equal to that for Ch^tia JNagpiir, as given above. G 2 Bengal. British. States. Northern 268 492 Western 240 — Eastern 388 86 North Bihar 489 — South Bihar 329 — Orissa 216 149 Chutia Nagpur 170 189 Assam. British. Siirma Valley Brahmaputra Valley Hill Tracts. 348 353 155 52 Bombay and Sindh. Bombay. British. States. Gujarath Konkan North Beccan South Deccan (Karnatak) Sindh 748 483 588 715 403 541 345 752 644 849 The four tracts included in the Bombay Presidency, excluding, of course, Sindh, how very different figures. The averages . of Grujarath and the South Deccan, or Karnatak, are both high, but in the former the aggregate is that of a generally homo- geneous community, on a single site, in the midst of a comparatively small area of irrigable land. In the latter, on the other hand, hamlets are common, since the black soil, though fertile, is not irrigable to any great extent, and thus necessitates the occu- pancy of larger areas for dry-crop cultiva- tion. The States falling within these two divisions show a different distribution. In Gujarath, the average is below that of British territory, by reason of the wide belt of forest-clad hills that is included in the Eastern States. In the South Deccan the villages are a little larger than in the surrounding British districts. In the northern portion of this division, which includes Khandesh, and the forests at the foot of the Ghdts, the States, which lie for the most part in the latter, show, as a rule, smaller villages. Along the coast, on the other hand, owing possibly to a different ti'eatment of hamlets, the reverse is the case. Madras'. The Madras average, on account of the very different tracts into which the Presidency is divided, is by no means a useful one. Allowing for the special con- stitution of the unit on the western coast, we find the largest villages on the Deccan plateau, where, as in the case of the Bombay portion of that tract, the congregation is due to the large holdings necessitated by the lightness of the soil, with the consequent growth of hamlets, subordinate to the main village. A third very distinct class of village is found in the Agency tracts of the east coast, which form a continuation of the hill-belt of Orissa and the Central Provinces. Omitting this, the provincial average rises to 692, the highest in India. This is probably the true light in which the distribution should be regarded", for throughout the rest of the province the average is remarkably above the general mean. The Malabar States have been already discussed in connection with the present subject, and the rest are so small that they need no special comment here. Burma. Burma comes last of the British provinces. It has not much of the village organisation of the rest of India, in which caste, from which this province is free, plays a considerable part in holding together the community. A trace of the village system is to be found, though slightly developed, in Upper Burma. The dis- tribution in the different divisions calls for no remark, save that it does not seem to follow at all the density of the population in general. Haidrabad. There remain the larger States and agencies about which a few words are required. Haidrabdd is returned in three linguistic divisions, the Mardthi, the Kanarese, and the Telugu- speaking tracts. There does not seem to be much difference between them so far as the village average is concerned, and the two first agree in this respect pretty closely. In the north-east corner of the last there is some hilly country that contains a number of small villages, but the average is raised by the size of those in the plains. iladras. British. States. North Coast Tracts 674 Agency Tracts 100 — ■ ^Northern Tracts 891 545 East Central 476 — West Central 708 — South Central 653 1,164 Southern 707 — West Coast 982 3,469 Burma. British. . r Northern 203 g_j Southern 286 o- ] Central 237 '^ Eastern 194 . r Arakan 172 1 1 Pegu o ] Irawadi 247 254 ^ |__Tenasserim 228 Haidrahad. State. Marathwada Karnatak Telingana 523 423 537 53 Mysore. State. Kashmer. State. Eastern Western 265 241 Kashmer Jammu Hill Tracts 221 320 354 Mysore presents no special features, except a generally small village. In Kashmer, the unit appar- ently varies in definition, so comparison is out of the question, different from each other. In Mysore and Kashmer. Eajputana. State. Western Southern Eastern 468 243 351 the and the Eajputana. Central India. State. Western Northern Eastern - 256 385 287 In Eajputana, the three divisions are very desert the villages are large, as they are in Sindh the lower Panjdb, and for the same reason. In south the Hill tracts are still in the occupation of wild forest tribes, who are averse from populous settlements, and so disperse by clans all over their territory. The Eastern States form the mean. They include more than half the population, and the extreme divergence between the other two gives them the preponderating influence in the general average. The Central India Agency is such a mixture of race, jurisdiction, and geo- Central graphical subdivisions, that the differences in the size of the village in the three great India. sections into which it is alone practicable to divide it may be purely arbitrary. The tract bordering on the Gangetic basin contains the largest villages, and Rewah, where the population lies thickest, comes second in order. The extreme west, like the south of Eajputana, is the refuge of the wilder tribes driven out of their former haunts in the valleys of the Tapti and Narmada, and their villages are little more than small collections of huts scattered over clearings in the forest. In a former part of this chapter the mean proximity of towns to each other was M6'*'i proxi- shown, and in the statement now under review, the corresponding figure will be found ™l*g^es for the villages. It may be worth while, perhaps, to explain that the proximity is a function of the square root of what Dr. Farr calls the areality of the unit selected. For example, in Sindh, where cultivation is sparse, each village, assuming equal distribution, is the centre of 12'17 square miles of country, and the average distance between it and any one of its neighbours is found to be about 3f miles. On the other hand, in Oudh, the villages are but 2f furlongs apart, and in the North- West Provinces, just over, and in Bengal just under, a mile. The three tracts last mentioned contain over 46 per cent, of the total number of villages, which are there so closely packed that they determine in great measure the general mean, which is a little under 1^ miles.* Table Gr. shows that in all other provinces this mean is exceeded. In the States, the average distance is greater, and. falls just short of two miles, but Haidrabad, Eajputana, Kashmer, and the Bombay States, all show a higher mean. The remainder of this table is given up to the proportional distribution of the places, whether town or village, according to population groups, and also of the population under the same groups. Nearly one half the total number of places contain less than 200 inhabitants. In British territory the proportion is a little lower, but in the States it rises above that figure. If we raise the limit to 1,000 inhabitants, it will be found thab 93 per cent, of the places, with 66 per cent, of the population, falls within it, and the admission of the smaller towns makes but an insignificant difference. The table shows how this ratio varies in the different Provinces and States. Omitting the exceptional cases of Ajmer and the Malabar States, nowhere does the proportion of population in villages of less than 1,000 inhabitants fall below 38 per cent., whilst in some, as in Assam, Burma and the small States of Bengal and the Central Provinces, it rises above 80, and even to 89, facts which speak for themselves when the rural element of the population is in question. House-room. In discussing the density of the population it is usual to treat, also, of the question of house-room, in connection with either the inhabitants or the area. In the one case, the mean number of persons per house is generally taken as the subject for comment. In the other, either the mean number of houses per square mile, or G 3 54 the mean area of which, on the assumption of equal distribution, each house is the centre. The proximity, too, which has been shown above to be a function of the areality, to use the nomenclature of Dr. Farr, is also in some cases a matter worth consideration. But it is with reference to urban tracts, principally, that these statistics are of importance, and to render the data complete, information is necessary regarding both the inhabited and the empty houses. In India, as we have seen, the conditions of town life affect but a small fraction of the people, and looking at the heterogeneous items of which a town in that country is often composed, and the large area of open land generally included within the municipal limits, it was held to be superfluous labour to do more than register the unoccupied buildings at the emimera- tion, without compiling the results. For the same reason, the areality of occupied houses is not a point, at present, of moment, except, perhaps, in two or three of the largest cities. On the other hand, it must be pointed out that where it is the custom to build the village in a compact mass, often surrounded by a high wall of masonry or hard mud, the sanitary conditions are deteriorated to the level of those in a large town, and if the commingling indoors, on intimate terms, of human beings and cattle, be an evil, the village is even at a disadvantage ; for the measures , of purification introduced into the former on the initiative of British administrators, is highly honoured in the breach in rural India. But the compact, or castellated, type of village is confined to the dryer tracts of India, and throughout Bengal, east of Bihar ; in Assam, Burma, the hill tracts, and along the coasts, the structure of the houses, as well as their arrangement on the village site, is quite different. The material is lighter, and the distance between the buildings is usually considerable, so that air circulates more freely than where the street, or continuous block of buildings, is the rule ; but the habits of the people are the same, and the drainage, especially in the comparatively level tracts, is very obviously deficient. The heavy rainfall seems to counteract a good deal. of the effect of this negligence, but when an epidemic has once got a lodging in one of these places, it is, no doubt, harder to dislodge than in the dry dirt of the north and centre of the country. ho°^ »" After it was decided to restrict the census inquiry to a return of the inhabited houses only, the usual difficulty arose as to the definition of a house. In England, after the matter had been discussed at an International Congress of Statists, the house was held to comprise by definition " all the space within the external and .party walls of the building." The committee of the aforesaid Congress that reported on the question, was unanimously adverse to the substitution of a proposed restriction of the term to what may be generally called a " tenement." In Scotland, and now in England, the latter term has been recognised, but subordinate to the general definition of the house. In India, the variety of structure with which the census has to deal is so great that it has been found impossible to frame a definition that would meet all the cases that would occur in the process ©f registration. In the Hill tracts one meets with, collections of leaf huts that are here to-day and gone to-morrow, because, perhaps, the genus loci has brought about the untimely death of some of the family poultry, or a forest-guard has threatened a visit of inspection to the adjacent reserve. In Haidrabad, on the other hand, there are the residences of the high officials, or connections of the Chief, covering several acres of ground, with numerous courtyards and detached buildings, into some of which none but a member of the household is permitted to go, and the whole is surrounded by what would be the " external wall " of the English definition. Between these, come numerous varieties of house or hut. First, the portable arrangement of matting and bamboo that is slung on a donkey by the vagrant classes, though sometimes stationary on the outskirts of a village for months together, according to the patience of the permanent inhabitants. Then, the more stable erection for the cultivator whilst engaged in watching his crops ; and so on, to the really permanent abode of the lower grades of village menials, with wattle and daub walls, which last for years, and a roofing of thatch or palmyra leaves, renewed as necessary, before each rainy season. The house, properly so called, is equally varied. In some parts of India it is the practice to wall in, with a thick hedge of thorn or rattan, a considerable space, over which the family expands in separate buildings as the sons marry ; but all is considered to be a single "house." In the north of India, where family exclusiveness is carried to the highest pitch, all well-to-do people adopt the walled enclosure, with a courtyard or yards, and a small room or verandah along the outer wall in which to receive visitors and stra,ngers. In Peshdwar, for instance, which is a city pre-eminent in this respect and, in a lesser degree in Lahore, there is street after street with hardly a single window in it. In the centre and south of India publicity is less disliked, and the 55 house, both, town and country, presents a less forbidding aspect. The material, too, is less solid, and, as a rule, the buildings seldom run to more than one story in the village, or two in the town. • Pitched roofs, tiled or thatched, are usual in the moister tracts ; flat-topped mud or brick buildings are almost universal in the dry plains of the Deccan and Upper India One great cause of this diversity, and probably the most important, is the variety Extremes of of climate that is experienced. The range of rainfall has received due attention in climate. the introductory chapter, but in connection with the housing of the people, the average range of temperature has also its value. From the following table it will be seen that in the drier parts of the country the range of temperature is considerably over 20 degrees, whereas the mean temperature is not excessive, as compared with that of the moister tracts. Table H.- — Showing the Temperature and its Monthly Eange. Eange of 20° F. and oyer. Range of less than 20° E. Station. Average Monthly Bange of Temperature. Mean Monthly Temperature. Station. Average Monthly Eange of Temperature. Mean Monthly Temperature. 'Dera Ismail Khan 28-3 74-3 'Cochin 12-5 80-3 Multan 26-7 76-3 Mangalore 12-6 78-6 Panjab - Ilfiwal Pindi Peshiwar 27-2 26-9 69-2 70-6 West Coast ■ Karwar Ealnagiri 13-5 14-5 78-6 79-2 Lahore 26-7 73-3 Bombay 10-8 79 '5 "Meerut 24-3 75-0 Karachi 17-2 77-2 Agra 23-4 77-8 'Negapatam 13-7 81 -.5 Ganges Valley 1 Lucknow 24-5 77-7 EastCoast■ Madras 15-9 81-8 Allahabad 23-3 77-8 ^Masulipatam 16-1 81-2 ^Benares 'Poona 23-0 23-5 77-8 75-8 r Calcutta Dacca 15-8 16-1 77-8 77-8 Deccan ■ ShoUpur Bellary 25-0 22-7 79-0 80-6 Bengal and Assam' Chittagong Dhiibri 14-8 14-8 76-3 74-2 _Karndl 'Khandwa 22-6 25-0 81-8 78-0 Sibsagar 16-3 72-3 Central Plains Jabalpur Nagpur 23-8 23-3 75-5 79-5 r Darbhanga Bihar < [Patna 17-0 19-7 76-7 77-3 ^Amraoti 23-6 78-2 'Akyab 14-0 78-5 'Haidrabad 24-7 79-3 Eangoon 16'4 78-8 Indus Jacobabad 30-3 78-7 Lower Burma "* Maulmein 16-0 78'5 Valley, &c. ' Bhiij 22-3 78-3 Mergui 15-6 78-5 Deeea 26-3 79-7 Toungu 19'8 78-1 In these last the range is very slight, especially along the sea coast. The result is that in the region of great extremes, in the north of India, the houses have to be built to keep out the burning heat of the summer and also the biting frost of the winter. In the Deccan, though the heat is great, the winter is mild, whilst the Central Plains furnish a very 'fair example of the medium between the two. Now, .along the coasts and in the Delta, even up to the Assam valley, the people enjoy a hot but equable climate, and have thus only the rain to guard against. But as soon as we get into Bihar, the extremes recede from each other, and a more substantial class of residence becomes desirable. Q 4 56 Social custom. Varying definitioii. The above is the principal physical cause for the diversity of building in India, and another is the material that happens to be most convenient. But, what with the climate that allows the people to spend most of their time out in the open air, and their main occupation, which compels them to so spend it, the house comes to be but a relatively insignificant item in the domestic comfort of the masses. In the dry season, as soon as it begins to get hot, shelter at night is considered in many parts of the country a troublesome and ridiculous excess, and the open spaces outside the houses are covered with cots or mattresses. In crowded streets, even, the footways are blocked by rows on rows of long white bundles, giving it the appearance in the moonlight of an exaggerated edition of the slab at the Morgue. ' But with the upper classes throughout the country, and with the middle also in the north, custom prohibits to the women any participation in this freedom, so the roof of the house has to be specially protected for their use, and in a large establishment, sometimes a small courtyard affords the means of taking the air, with such exercise as is held to be necessary for the fair sex. Social custom, too, has a good deal to do with the construction and arrangement of the dwellings of the people in more ways than the above. The relative prevalence, for instance, of what is known as " joint " or " divided " family life amongst the Brahmanic classes, that is, the bulk of the population of India proper, determines the partition of the house inside, or even the number of doors. Then again, there are numerous degrees of detachment. If the link is only severed so far as property is con- cerned, the family often continues to eat together, and to cook at the same fireplace. But it may be deemed necessary to reverse the operation, and then the property remains in its integrity, but the family adds to its fireplaces, and eats in separate messes. The due arrangement of these fireplaces, too, has to be considered, as there is risk of offence to the supernatural if the mouth does not open towards a certain point of the compass. In many parts of the country, in the present day, it is said to be the custom for the younger branches to build themselves entirely separate houses, and to leave the paternal nest altogether, on which matter the provincial census Superintendents have in some instances found evidence in the returns. Amongst certain classes, again, chiefly in the Central Belt of hills, there is a general refectory or meeting-roonl for the whole unmarried population, and 'the spinsters are sent off after the evening meal to a dormitory at one end of the village, and the bachelors to one at the other, the houses of the Benedicts of the community inter- vening. The results of this arrangement are not said to be different from those of any other. Upon considering all this variety, and the difficulties in getting a uniform inter- pretation of a general definition that- arose in 1881, it was decided to prescribe a standard on which the provincial census Superintendents were to engraft special additions to suit the circumstances of the different parts of their charge. It was the object to thus obtain a return that would be uniform for the province or State, though it might be inconsistent with that of other parts of India. Clearlv, therefore, the subject had to fall out of the ranks of the Imperial series. So far as the latter is concerned, the chief aim was administrative ; namely, to show the enumerator what buildings he had to visit. It has, however, been treated provincially in all the reports, so that a certain amount of useful information has been recorded on it. It is not proposed to review the return generally, though this work would be incomplete without the quotation of the' figures for the different provinces and States that have compiled them. For the sake of comparison, the corresponding return of 1881 has been added. The reduction of the average number of persons per house from 5-8 to 5-4 is due to the changes in Bengal and the ISTorth-West Provinces, and these, in turn, are clearly attributable in great part to a change in definition. In the last-named province there are numerous cases in which a courtyard surrounded by many separate tenements opening on to it is entered by a single gateway. These were formerly treated as one building, but on this occasion they were enumerated by tenements. In Bengal, again, the test on this occasion was the custom as regards eating together, or what the Provincial Super- intendent calls by the convenient name of the commensal test. Each separate cookins; place was counted as a house wherever the building was obviously outside the standard definition. In the city of Bombay, too, it is plain that the enumeration must have been by tenements on this occasion, probably counting as such those which had an independent entrance from some space either public or common to all the residents of the building in question. For further details the reader is referred to the provincial reports on the census. 57 Table J. — Showing the Average Number of Persons per House. Province, &o. 1891. 1881. State, &c. 1891, 1881. Ajm^r 5-2 5-3 Haidrabad •■ 51 5-3 Assam 4-8 5-4 Baroda - ' 1' 4-5 4'6 Bengal 5-4 6-3 Mysore 5-5 5-7 Berar 4-9 5-7 Eashmer 5-7 — 'Bombay 5-6 5-6 Bajput&na 5'5 4'9 . Sindh 5-6 5-5 Central India 5-3 5-6 ■ Upper Burma 5-3 -^ Bombay States 5-0 5-1 , .Lower Burma 5-4 5-5 Madras States 5-1 4-9 Central PrOviiioes 5-0 4-2 Central Province States 5-3 4-6 Coorg 7-0 7-9 Bengal Province States 5-3 5-6 Madras 5-3 5-5 North-West Provinces States 6-0 5-8 North- West Provinces 5-8 6-8 Panjab States 5-8 5-9 Oudh Panjab 5-5 6-9 5-5 6-7 Madras City 7-5 8-4 Calcutta Bombay City 9-6 14-4 11-6 26-3 Bangoon Town India 6-4 6-4 6-5 5-8 . / 78388, H m CHAPTER III. , The Movement of the Population. The true greatness of a State consistetli essentially in populatioii and breed of men.— Bacon. In the foregoing part of this work the population has been regarded as stationary, that is, as it existed at the time of the consus was taken. It has now to be considered as in motion, in its relation, that is, to what it was 10 years ago, on the one hand, and, on the other, to the general change that, judging from recorded experience, is likely to have taken place in it by the time the next statement of account is obtained, at the beginning of the 20th century. General In this, its dynamical aspect, a population is the resultant of certain opposing conditions, forces, whose combined action tends ultimately to equilibrium. In the present stage, however, of the world's development, there is an almost rhythmic variation in the intensity of each, effecting continuous change, the bias of which is just now in favour of the positive, as distinquished from the preventive, elements. There are various ways of grouping these elements, and, on broad lines, we may adopt the two classes of the physical, or domestic, and the political, or foreign. IFnder the former head come the natural forces of birth and death ; in the latter we place war and migration. Underneath these general combinations, again, may be distinguished influences that are practically constant from those which are accidental or temporary. The reproduc- tive instinct, for example, is constant, but its results are stimulated or repressed by artificial considerations, such as, amongst others, the prevailing views on marriage or inheritance. It may even be qualified, some hold, by a modification of type in a community which has reached a certain stage of development, in which case the restriction borders on the permanent. Death, again, is a debt due by all, though most of us are like FalstafF, loth to pay it before its time; but is liable to receive the fortuitous assistance of war, pestilence and famine. The change, too, resulting from migration may be accidental, or else part of the ordinary process of internal expansion. The conclusions arrived at by Malthuson this important subject have been severely criticised, and often misquoted or misapprehended, but, in the main, they have not been disproved. In certain particulars, no doubt, his deductions were too sweeping, and in others, modern science, in the progress of physiological investigation and in the experience drawn from a wider field of observation, has had to introduce qualifications derived from sources of which Malthus was ignorant. For instance, in estimating the means of subsistence, he seems to have depreciated the results of an increased and improved application of human industry to the production of food from natural sources, and to have laid too little stress on the power of dispersion in sustaining any particular population. His view, too, that the power of multiplication was only restrained by what he termed the preventive checks, appears to be controverted by what we know has happened since he wrote. For, allowing that the proportion of married women of the reproductive age is the main factor in the natural increase of the population, we find that instead of the maximum number of children that could be produced during the 20 years included, as a rule, in this period, the average, even under highly favourable conditions, physical and moral, is about four per marriage' and only six at the maximum. We see, also, that in countries where every circum- stance is in favour of multiplication, unrestricted by either postponement of marriage or abstinence from it, by war, starvation or disease, and stimulated, perhaps, by the immigration of adults of both sexes, the rate of increase is considerably below that which might, ex hypothesi, be possible. There is every reason to suppose, accordingly, that Malthus' view ignores some important physiological sequence that has not even yet been ascertained. We have, however, the suggestion, to call it by the least assumptive title,_of Mr. Herbert Spencer, that pressure of population consequent on excessive fecundity, tends to provide, in some degree, its own corrective. For this pressure implies an increased demand upon the intellectual and nervous qualities of life to meet it, and the stock of nutriment being the same, these latter can only be satisfied at the expense of the reproductive faculties. The more the former are brought into play, the less there is left for the sustenance of the others. Life may be, and probablv is, lengthened, but the community is neither so numerous or so physicallv well equipped as the well fed savage. If this doctrine be correct, and it has much to recommend it, one of its consequences may fitly be mentioned here. It is, that according to the theory, every community has to 'pass through a stage of what we may caU 69 probation. Between races and even the larger political entities, sucli as nations, a higher, type of civilisation in the one and of character in the other, is pretty sure to prevail, ,^nd this is apparently what Bacon implied in the latter portion of the phrase quoted at the .head of this chapter, and ^hat one of our greatest modern statists, Dr. Farr, meant wh'en he said that the " character of every race pf men, " allo'vying for accidents of position and time, is the real limit to its numbers in the " worild." But in the case of smaller aggregates than the above, where, that is, competition lies within iiarrowed limits and is actuated, by more individualised impulse, intelligence and what are called the social virtues, must be, for a time, and a considerable one, at a discount. For, by the hypothesis, they result in diminished fecundity, so that, until they have permeated widely , through the community, the class which exercises them must inevitably fade into numerical insignificance relatively to the reckless, prolificity of those who have nothing to lose by the indulgence of their strongest instinct. This is an aspect of the population question which it has luckily not been necessary to take into consideration hitherto in connection with India, but the peculia?, social organisation of the great mass of the people of that country renders it impossible for enlightenment to progress at anything like the rate of its numerical growth- The period of probation, therefore, during which these multitudes will have to be passing through the discipline of hardship before they are mentally adapted to their new circumstances, must be a peculiarly long and trying one. Luckily it seems to be still far off, and the approach to it is by steps almost imperceptible. For before pressure of the character above described can be felt by a community it is necessary tha;t dispersion should have reached its limit. By dispersion is not here meant solely the migrations of a single community, but the corresponding movements of all communities to tracts which allow of their pro- viding food to supply the wants of others, over and above the stock raised for their own maintenance. It is the facility of this transfer of the results of labour together with the greater industry and intelligence brought to bear on the soil where the arable area is fully occupied, that makes it possible to say- that in the present day the food supph' has increased not only in line with, but even in advance of, the growth of the population, and, other things being equal, this will continue to be the case, so long as there is land ^o be found on the globe fit for remunerative tillage. We can now consider the part ' respectively played in the movement, of the Conditions population of India by the birth-rate, by famine, by epidemic disease and other in India, causes of death, and by migration. War has happily been, for this generation, a factor in abeyance. - Of the rest, the natural causes of change, that is, the birth and the normal death rates, are necessarily entitled to the first place in the list, and in connection with thein, the special causes of mortality, by epidemic and famine, will have to be dtecussed. ■ Migration will then be treated separately^ NoW, the nuniber of births depends, as was reinarked, above,. very much on the Marriage proport;ion of mamed women of the reproductive age, so we are here brought face customs. to face with the marriage customs of the people and the tendency towards matrimony prevailing in India. In anticipation of a more detailed examination of this wide subject J which will be specially handled in a later chapter, a few general observations must suffice to explain its connection with our present topic. We should omit Burma, for the present, from the question, since the social system of that province is entirely different from that of ,the rest of the country. In India, as thus restricted, marriage is regulated by a few leading principles which nominally prevail amongst at least three-fourths of the population. Strict conformity with all is probably confined to a not very large minority, but such is the pervading influence of the system indigenous to the country, that a good many of its special tenets have been adopted by most of those communities that are brought into close contact with , it, even though the orthodox prescriptioris of their own system may be opposed to them. One of the most prominent universal duties inaposed by this heterogenous public opinion is that of marriage. In the case of the Brahmanic community this is intelligible, as their current system dooms a man to a particular region of pandemonium unless he leave a son to perform the proper obsequies for his relea,se. The duty is also connected with the law of Inheritance which is, to some extent, binding on the masses. The forest tribes conform as they become civilised and are incorporated with the rest. The Musalmdns, who. are, for the most part, converts from the ranks of Brahmanism, have not' abandoned it,* and the rest of the denominations of Indian growth equally observe, the obUgation. Then, again, amongst the higher and middle classes of all denominations there is the"" general feeling that the paternal hearth is disgraced by H 2 60 the presence of a girl who has arrived at womanhood unmarried. Polygyny, too, though undoubtedly rare, is allowed to all, whilst polyandry is confined to two or three comparatively small communities. It is scarcely necessary to dwell on the mam fact indicated by these tendencies, that is, the enormous number of births that is likely to result from them. At the same time, it must be pointed out that it is the universality of marriage, not the early age at which it is contracted, that is here in question. The latter has sometimes been erroneously put forward as one of the causes of an increase in the child population, whereas its chief and direct effect in India seems to be only to shorten the mean lifetime of a generation, a,nd not to increase the number of births per marriage, possibly owing to the physiological cause mentioned above, the nature of which has. not yet been ascertained. Furthermore, the relative age at marriage is largely a matter of local custom, as well as of social position, and in certain tracts the upper classes marry, as a rule, at a higher age than the lowest in others. It should also be borne in mind in regard to this part of the subject, that the term marriage includes both the formal betrothal, which is the principal ceremony, and the subsequent handing over the bride in manum viri, the interval between the two being, as above mentioned, dependent on the local custom or the rank of the con- tracting parties, as the case may be. Amongst the Hill tribes and many of the labouring classes, the ceremonies are not distinguished, as girls are not married in either sense till they are grown up. In its ceremonial aspect, the betrothal is the binding and more important part of marriage, but from our present standpoint it is superfluous to pay attention to more than the later stage of this social arrangement. Thus we find that 94 per cent, of the unmarried females are below 15 years of age, and that though the proportion of women between 15 and 40 years of age is but a trifle in excess of the corresponding ratio in European countries, in India, no less than 84 per cent, of them are married, whereas in Europe, if we omit Hungary, where the ratio rises to more than 63, the average is not above 40.* The period in question, too, contains in India about 70 per cent, of the total married female population, but in Europe, excluding Hungary, as before, the average is only just above one half. This fact takes us a step further in the analysis of the return, for if we sub-divide the period into two, it seems that of women in India between 15 and 25 years old 87 per cent, are married; but in Europe the highest proportion, to the west of the Leith, is in France, where it is only 22. In the remaining period, from 25 to 40, the ratio of wives in India falls to 81 per cent., whereas in the West, it advances to about 70. The divergence is here due to the Indian custom as to widow-marriage. The orthodox Brahmanic view, whether it be well founded or not we need not inquire, absolutely prohibits the marriage of females whose husbands are dead, so that a girl betrothed in infancy, whose affianced happens to die before the latter part of the ceremony has taken place, is as much a widow as she who enters that* condition at the decline of life. This rule is only strictly observed amongst the comparatively small communities of the upper classes, and throughout the masses there is great variety of practice. The lower and lower middle classes are, on the whole, influenced chiefly by local opinion on the matter, and where adult marriage is the rule, there is generally found no objection to the " maimed rites " and less expensive procedure involved in bringing home a housewife at second-hand. On the other hand, the general tendency of socisil aspirants to adopt the customs of those immediately above them, is extended to the case of the widow of course, and when a man of this class has, as the French say, " arrived," he makes the fact apparent by secluding his womenkind and abjuring widow-marriage. The above remarks apply to the Brahmanic community only, or, at most, to the fringe of Hill tribes that border on the plains. The Musalmans have no prejudice in the matter, teste the first matrimonial venture of the Prophet himself, but even amongst them the pervading atmosphere of Brahmanism has had a certain efiect. On the whole, the proportion of women who are returned as widows is 17"6 per cent, of the total of the sex, or about 3-5 to every widower. In Europe, the average ratio is about one half the above, and there are only about 106 widows to 100 men in the same condition. To some extent, therefore, we may take into consideration this rejection of second marriages amongst females as a check on the number of children that might be brought into the world under the operation of the great stimulus given by the religious sense of the majority to the indulgence of the strongest animal instinct of humanity. But, looking at the prevalence of marriage, it is clear that more than tbe existence of a few millions of widows more * The more correct period for Indian women would be from 13 to 35 years of age, but for the sake of comparison the higher limitation has been adopted. 61 or less, is required - to account for the comparatively slo-w growth of the population under the impetus of so enormous a number of births. The clue is to be found in the accompanying high, mortality. The birth-rate is indeed very far above that of Approximate any European country, if we except Russia, and reaches nearly 48 per mille on thd birth-rate. whole country. But the death rate is equally abnormal, even if we omit the more frequent occurrence of famine and epidemic disease in India, and may be taken to reach, on an average, 41 per mille. This matter will be discussed in its due sequence, but, speaking generally, it can be attributed to an excessive mortality amongst infants of both sexes, a large mortality amongst young mothers in childbirth, and a general want of " staying power " after about 50 years of age. Of these factors, the first is the most prominent, as about 26 per cent of the children born do not live to the end of the first year of their life. In England the rate is not higher than 15'6. The operation of the second cause is more difficult to determine, as the return is complicated by the admitted repugnance in some parts of the country and amongst certain classes to communicating any information about the married or nubile women of their household. As to the last, the returns as they stand uncorrected indicate more than five per cent, deficiency of persons of 40 years old and upwards, as compared with the age-tables of Europe. The general divergence of the constitution and age distribution of the two communities can be appreciated to some extent from the average duration of life, which is little more than 24 years in India, against nearly 44 in England. These figures are given merely for the purpose of comparison. They do not show the actual mean age of the population, as it stood at the. census, but the higher estimate obtained by considering the community as stationary. The .former averages would, necessarily, in the case of increasing populations be some years lower. StUl, the fact is made prominent by them that in India the generations must succeed each other with much greater rapidity than they do in the West. In estimating the growth of the population of India by what we have termed Birth and the natural increase, that is, by the excess of births over deaths, we are without the '^^^^ ^^S^^' aid of complete registers of these occurrences. Registration has been in force for a ''*^°'^- considerable time, and a good deal of attention has been paid to it, but the circum- stances of the country are against it. In municipalities, especially where town-duties are levied on imports, the deaths are registered with approximate accuracy, as the corpse has to be carried out to the cemetery or burning-place past one of the toll-bars. Births, however, even here escape registration to a very considerable extent. In the rural tracts, where the country is open, the villages close together, and their staff" of officials fairly complete, the results are, on the whole, very good ; the births, indeed, are probably as correctly entered as in the towns. But in the hill tracts, where locomotion is difficult and the villages far apart, even if stationary, and ' the people not only illiterate, but suspicious of everyone who is not equally so, the returns are too often altogether neglected, or kept by the headman by means of knots in a piece : of string hung up to a beam in his roof till the next visit of the accountant or other officer, who writes up the register from dictation. In many parts of Bengal, too, the absence of a village staff" placed registration for a long time in the hands of the lower grades of the rural police, so that the return was trustworthy in towns only. Then again, there are tracts where registration is not even attempted, owing partly to . the density of the forest and the sparsity of the inhabitants, or again, to the nature of the tenure, which varies greatly in its degree of independence of State control. The large area under the feudatory Chiefs is excluded from the returns. In a few cases, . such as Mysore and Baroda, there is a system of registration, but in most of the larger States information of this sort is not available. Finally, . we come to the personal grounds of. objection to the system. The householder of rank dislikes inquiries of any sort as to .his family aff"airs. The rest do not fall in with the official idea of its utility, even if they understand the question. The death of an adult male member of a family is an event known to the whole village, and is registered accordingly, so is the birth or the death of a son. But mothers, wives, and daughters pass away without leaving a ripple on the surface of village life, and it is with respect to them that registration is most deficient. It is plain, however, from the returns, that greater accuracy is annually being obtained, and in one or two provinces, as was remarked above, the results for the 10 years that elapsed between the two enumerations so nearly coincided with those of the census that both operations may be congratu- lated on their successful issue. Elsewhere, the latter is the only means of, as .it were, taking stock of the population. Even in England, where registration of domestic occurrences has reached so high a degree of accuracy, the last census shelved that the results, apart from migration, were not quite what were expected. There is H 3 62 Epidemic disease. Cholera and small-pox. Fevers. anotlier point in connection with these returns for . India that requires notice. The increasing accuracy of the registration, especially where the latter has been hitherto backward, is likely, if not duly discounted in the returns, to lead to the assumption that changes are taking place m the birth and death rates which, in fact, are not taking place. This is the more noticeable when, the population on which , the rates are calculated is maintained, as it often is, at the figure of the I^sfe, census throughout the decade, so that even though the registration may have been uniform in its degree of accuracy during the whole interval, there is a sudden descent in the rates when the increase of perhaps 10 or more per cent, in the population, shown by the census, is brought into the returns. With all their deficiencies in respect , to, .the absolute figures, these returns afford most useful data in comparing one year with another; for the error to be found in them is in all probability constant and ui3.iform throughout, the whole series. It is in this capacity that they have been use^^ in preparing the corrections to be applied to the age-returns, which, as will be showTj in a later chapter, enable us to dispense, to some extent, with the actual numbers. Having thus disposed of the main normal factors in the natural increase of the population, it is necessary to refer to those which we may call incidental, the first of which is the check on the increase produced by the prevalence, in annually varying proportionate strength, of certain epidemic diseases. It is not easy to treat this subject otherwise than very generally, for the diagnosis of the village accountant or the local constable is very liable to error, and except in the hospitals and dispensaries, the classi- fication of the causes of death leaves much to be desired. Small-pox and cholera, however, if, indeed we may call them epidemics in India where they are always present, are probably more correctly registered than most other fatal maladies. The symptoms are too well known, and the disease, too, in both cases, is under the special control of a certain female divinity, who is to be propitiated only through the mediation of the classes descended from the pre-Aryan population. The rest of the ills to which Indian flesh is heir to, excepting accidents and snake-bite, mostly come on to the returns under the generic title of fever. Taking the return for what it is worth, we have had, during the past 10 years, a population under observation averaging about 197j^ millions, with a mean annual number of deaths amounting to 5,140,000, which seems to indicate an omission of at least one in three. Of those registered, the 10 years' average- includes about 309,000 deaths from cholera, yearly, with the maximum of 475,600 ; 126,750, from small-pox, the highest number being 333,380, and 3,397,300 from fever, with the coiresponding limit of 4,110,000. Of the remainder, a number just short of a thousand is unclassed ; and accidents and what are grouped under the head of bowel complaints, account for the rest. Thus, to fever are attributed Q6 per cent, of the deaths, to cholera, 6 ; :to small -pox, 2 ; to bowel complaints, 5 ; and 21 to injuries and unclassified causes. The latter include, roughly speaking, 60,000 accidents, of which a considerable proportion are due to drowning, and 20,000 deaths from snake-bite. Cholera and small-pox are the two main causes of abnormal mortality in India, apart from famine and certain special outbreaks of fever, which will be noticed below. Not a year passes without cholera in some part or another of the country, and there seems to be no sign of its becoming rarer or less fatal. Whatever may be its origin, its dissemination is no doubt largely due to the immense congregations of pilgrims at certain seasons of the year, especially about the hottest time, to bathe and drink at one of the many sacred rivers or pools of the country, just about the month when the water is at its lowest. In spite of all the sanitary precautions adopted, the outbreak is still a matter of chance, and once it happens' there is no limit to its local extension. As to smail-pox, though it cannot be said to have been stopped by the greater prevalence of vaccination nowadays, it is said to be of a milder type in' some parts of the country where it was formerly frequent and severe. The returns of blindness seem to indicate this to a slight extent. ; Fever, as has just been remarked, includes a variety of diseases, amongst others, influenza, in the form in which it was prevalent during the past three years. There are, however, certain classes of fever that seem confined to special localities, which they ravage for a few years, and often disappear as unexpectedly as they broke out For instance, in the Brahmaputra valley of Assam, the " black sickness " (kala azar) that broke out some years since, has been peculiarly destructive to life along the southern bank, and has also crept across to a few tracts on the northern. For some time it baffled medical research, but its nature was thoroughly investigated in 1890 by a competent expert, who found the disease to be largely due to the insanitary habits of the villagers. The name he proposed for it was parasitic ansemia or 63 anchylostomiasis. Whatever it may be, its results are painfully apparent in tlie two districts where it has been rife for the longest period, and the Provincials Super- intendent of the Census attributes to it a loss of over 100,000 people during the decade. • Another instance of epidemic fever is that popularly known as the Bard^dn outbreak, from the name of the district where it was specially prevalent some years ago'. The' tract, however, has obtained this bad eminence unjustly, for it seems that) the: disease originated further in the delta, about Midnapur, where it is attributed to the water-logging consequent on the choking of the natural drainage channels of this part of Bengal,- by reason of the gradual changes in the course of the main estuaries. This process has been in operation, too, high up the delta, even to the border; of the submontane tract, and affects the returns of four large districts. It is not only in the lower part of the Gangetic basin that water-logging has occurred. In the south-east of the Panjab the natural drainage has been obstructed to some extent, and portions of the Karnal and Dehli district have passed out of popular favour^ apparently for good sanitary reasons. Remedies have been tried, and others are about to be applied < to the condition of the tract by Government. The case of the southern portion of tke Ganges Doab was mentioned in the last chapter. Here, however, there does- not seem to, have been so much an increase of mortality as the •abandonment of the soil because it deteriorated for agricultural purposes. Along the borders of the Tarai, or sub-Himalayan forest and grazing tract in Rohilkhand on the other hand, fever has increased in prevalence during the last decade or so, though it is not said to be of so special a type as that of Bard wan or Assam. On this point the report of the Provincial Census Superintendent, which has not yet been submitted, will no doubt contain full comment. In other parts of the country there have been outbreaks of fever due to some local cause, such as that in Amritsar in the Panjab, where the city population fell off by 11 per cent., whilst the rural tracts surrounding it continued to increase. In other cities, too, the malady we now call influenza grew to the intensity of an epidemic, and carried off numbers of tjie inhabi- tants in a few weeks. These examples, though not covering the whole field, suffice to show that in India the abnormal influences affecting the death rate are by no means as rare as the equable climate; and the healthy outdoor life of the population at large would lead one to suppose they would be. Tke next of the influences that we have to consider is that of famine, with which Famine. we have in India always to reckon. Most fortunately, the 10 years under review have been almost free from this, calamity, and the one or two cases of serious failure of crops that did occur were purely local and restricted to very narrow limits, both territorially and with respect to. the population affected. In fact, the only occurrence of this description worth mentioning is' the scarcity that prevailed in the n,orthern portion of the east coast of Madras, in 1889, and even here the direct effects :were compai'a.tively .small. The season following the census, however, was unluckily marked by distress in the southrcast Deccan, as well as in Eajputana, and by scarcity of a milder type in Bihar, but this is a period falling beyond the limits of the present subject. It is otherwise withi, the great famine of 1876-78 in the Deccan aind South India, which has impressed itself rudely on the census returns.. Here, as' in the case of Orissa, in 1886, and Rajputana, two years later, and again, of the North- West Provinces, in 1861, the effects, will be marked out in the age-tables until the generation that suffered them has passed out of life. But for the present, we have only to consider famine as one of the checks on the growth of the population, not in its detailed action on the latter. That check is exercised in a twofold manner, directly and indirectly. It not only increases the number of deaths, but it tends to diminish that of births otherwise than by merely destroying possible parents, as was the case in Thebes of old, oure yaq eKyava ■ K'Kvrd.g ^S6vog au^erai oiire tokoktiv 'Irjioav KQCfjLaTwv avep^ootrJ yuvMKeg. As regards the first, the number of people who die from actual want of food is probably small conipared to the deaths which result from the greater hold which disease gets on those who are enfeebled by diminution of their usual supply of nutriment,' ' ^Chus, in times of scarcity, the mortality from ordinary causes, such as bowel complaints and interinittent fever, rises considerably above the normal rate, since many succumb who would in. ordinary times offer a successful resistance. The second of the results just mentioned was very prominent in the age returns at the census of 1881 for the .X)ecca,ii and Southern India, and reappears at the age of 10 to 14 in those of 1891. From these data it is clear that famine is most felt in the first four or five years of , life It then seems to pass lightly ovei* the adults, and to fix' ott the agqd, but only H'4 64 Probable diminution of normal mortality. Males. Females. Famine. Normal. Famine. Normal. ["Total increase Ma DBAS < Increase of children l_ under 5. r Total increase Bombay < Increase of children 1. under 5. 20-56 72-83 20 03 66-34 11-20 19-69 10-05 15-87 19-58 71-82 19-11 66-60 12-10 19-14 10-63 16-97 ■where the distress is acute, for the evidence on this point is hot conclusive, as it is on the other. Now, as the reproductive ages are the least affected, one would expect to find the process of replenishing the gaps in the depleted population in full operation within a year of the return of normal prosperity. But even where the famine was followed by a bumper harvest this did not happen, and the cause of the check in reproduction was undoubtedly physical weakness, resulting from the long spell of insufficient nourishment endured by the masses. It took from three to four years, according to the returns, to restore the vitality of the worst tracts, but the marginal table shows that since then lABLE A. a, considerable amount of lee-way has been made up. In regard to this table, it should be borne in mind that the famine was far more severe and of longer dura- tion in Madras than in Bom- bay. There is also an in- dication that the resistance offered to distress is greater on the part of females than of males. On the whole, we have only to read of the terrible mortality that accompanied the historic famines of 100 or 150 years ago to a'ppreciate the advance that has been made within the present generation in administrative experience of how to deal with these calamities, so that their effects may be mitigated, though to prevent them is impracticable. By the aid of lines of railway and telegraph, the tract threatened with scarcity is brought into speedy communication with those in which the harvest has been abundant, so that from the latter food pours in as long as the prices at the markets of the former keep the transport remunerative. The case of the classes thrown out of work by a failure of the crops, who have neither grain of their own nor cash in hand to buy what is physically within their reach, has been duly considered, and so far as a crisis of this sort is concerned, the problem of State aid for the unemployed has been solved. An estimate based on local inquiry, collated with the census returns of caste and occupation is prescribed for each district, showing the approximate numbers of the community that are most prone to suffer, classified according to the order in which they are respectively likely to be thrown upon State support. Estimates and plans of under- takings of public utility involving the employment of the requisite amount of unskilled labour are kept in readiness by the Public Works Branch of the Local Grovernment, and by each of the latter a code of rules regarding famine administration has been framed, which includes minutiae of sanitary arrangements, daily tasks, the lading out of camps of work and refuge, with other measures of relief that the experience of the last 30 years has proved to be the most efficient in the circumstances of the tract and population to which they are to be made applicable. The loss of life which inevitable in time of famine is thus likely to be reduced to a minimum. The present system of administration is conducive also to the preservation of life in less abnormal circumstances. The spread of vaccination, though uneven, is doing much to mitigate the ravages of small-pox. Cholera, which it seems impossible to prevent altogether, is localised by segregation, or by the strong measure of prohibiting religious gatherings, whenever they are likely to lead to an outbreak of this scourge, and in all such cases the sanitary arrangements of the locality are placed under the control of special Superintendents. As for normal disease, every year sees an increase in the number of dispensaries, which are, in fact, small hospitals under trained men, scattered about the rural tracts, whilst in larger towns the lower grade of medical practitioner turned out by the Universities, is growing in popular favour against the rivalry of the herbalist and exerciser. There seems, again, no immediate prospect of any general amelioration of the views of the populace respecting marriage, which wiU contmue, therefore, to be one of the chief duties of man, whilst the extension of trained obstetric aid of late years will in time probably do something to counteract the results on immature mothers and their progeny of the ruthless empiricism of the barber's wife, who at present is the chief officiator in this capacity. But can there doubt as to the result towards which all these provisions tend ? In countries where there are no large tracts of unoccupied arable land the rate of increase of the population is, other things being equal, inversely as its numerical growth. Every census shows that in compliance with the tendency noted in an earlier portion of this chapter, as a country fills up, the annual rate of increase diminishes IS be any 65 The conditions on "wliich that rate can be maintained at a uniform level are either the increase of return from the cultivated land, to which the limit is soon reached, or dispersion of one of the two kinds already specified. The community must betake itself to occupations other than food-producing, and indent for their nutriment on other parts of the world, where the products of their industry are in higher demand than food ; or the dispersion must be over fresh land in their own country. Whilst and where this last is to be got, pressure of population amounts to no more than local congestion, for so long as there is room, actually or potentially productive, relief is only a matter of time, more or less according to the character of the people concerned, their adaptability to change, and the nature of the tract within their reach. Now, in India, there are several tracts where the agricultural skill and experience Quality of of the people have been said by experts to get the maximum yield out of the soil. In agriculture, others, the harvest - does not respond so thoroughly to the care bestowed on it owing to the scanty or uncertain rainfall, but the aid of irrigation can he called in to preserve the crops. In most parts of the country there is a great waste of the material available for fertilisation. Cattle are very numerous, but they are turned out to stray over waste land, or where they can pick up a little grazing more or less scanty, so that their droppings are either wasted or collected only for fuel, and their bones are being exported, for use in Europe in increasijig quantities. In some parts of the country, sheep and goats are systematically hired out to be penned at night on the field, but this practice can affect but an insignificant portion of the dry-cropped area. As to rotation, experts have a good deal to say against the somewhat primitive cycle received by the Indian husbandman from his forefathers, and it may, perhaps, be possible to increase or sustain the productiveness of the soil by a change in system, but this, like so many other suggestions in the same direction, is at present a matter of conjecture. The variety of character amongst the cultivating classes, too, is a factor in the situation, as well as that of the soil. The chance, under the present conditions, of many more tracts in India rivalling the fertility of Oudh and the great deltas is about equal to that of the scratcher of the hillside in the forests, or the many-acred driver of the Maratha plough, settling down into a market gardener. So far as lies within our present purview, neither one nor the other is probable or impossible. It appears from what has been sketched above, that throughout the greater Prospects of portion of India the occupied land has, probably, not yet reached the limit of its dispersion, productiveness, and in the preceding chapter it was shown that congestion of the population was at present found to comparatively a small extent. We have to consider now the question of dispersion, whether vertically, by change of occupation or produce, or laterally, by change of locality. As to the former, circum- stances seem to point to a long continuance of the present position of India as a self-sustaining country. With its existing population it manages to set aside very considerable areas for the production of raw material other than food, and throughout the decade under review, the outturn of this class of produce seems to have increased rather than to have been restricted by the growth of the population. From one quarter, cotton and wheat, from another, indigo, jute and tea ; rice from a third, and oilseeds and opium impartially, pour into the seaports, in close response, apparently, to the opportunities offered in foreign markets. The easier task for the cultivator would doubtless be to confine his efforts to the growth of food-grain or pulse, so his free adoption of the alternative proves that he appreciates its superior advantages. As for manufacturing industry, its progress, though rapid of late, has not been diverse enough, either in kind or locality, to make much impression on the country at large. It has certainly hitherto won only a comparatively small portion of the community away from the agriculture to which the letter is traditionally devoted. For one thing, India is at present at a disadvantage in comparison with many of its compeers as regards a sufficient native supply of good coal and iron. Its labour is plentiful enough ; whether it is cheap or not depends on its quality relatively to that of the niore highly paid wage- receiving class elsewhere. Some branches of industry, it is true, have taken root to an extent that seems to open out the way for the transfer of capital to India from countries where, for various reasons not yet prevalent in the East, profits are, to. put it mildly, uncertain. Amongst these are, of course, cotton and jute works, to which ,may now be added paper-making and articles of leather. The wider extension of any of these operates towards the dispersion of the population from the land, and so far tends to lighten the burden as it becomes heavier in the course of years. Finally, there is th^ question of lateral extension. To measure the possibilities in this dii?ectioTi it is necessary to ascertain the area pf arable land still available foif f 75388, I 66 Migration. Immi- gration. cultivation, and this, as was shown in a previous chapter, we are unable to do,, with even approximate acduraicy, for, the country at large. It will be more convenient, therefore, to touch briefly -upon the point in connection with the provinces for which the best-information is availa-ble. For our present purpose, the returns . of birthplace, which form part of the general series, will enable us to appreciate to; some extent the tendency towards migration within India. This tendency, it is clear from the tables, is at present remarkably weak ; and that this should be the • case is no more than is to be expected from- a -commimity so devotedly agricultural, a class invariably opposed-— •; , , .Migrare yetusto , , . De nemore, et proavis habitatas linquere silvas. In India, too, there is the' additional attractive power of the village organisation and the prevalent connubial system. With regard to expansion within the limits of the land appertaining to the village (itself there is no difficulty, but cultivators from outside: are, as a rule, only admitted- with considerable reluctance, and in many' parts e£ India 'would not obMnland at all. Then, too^ wherfe) the available waste land lies ii^ a compact tract within an easy distance of the ancestiral village, the populatipa is able to advance on it, as, it were incline, and is thus in a position to form fresh communities on exactly, the old model, and to keep in touch also witk the home' and tradition of their forefathers. The objection to. . agricultural migration "begins where the tract to be occupied: is separated from the, native land of" the emigrants by a population differing from the latter in race or language, and' except in a few special cases, attempts at migration of this sdrtr, to any great extent, have not yet been successful. In Assam,, a good' m^ny of. the immigrant labourers on tea plantations have settled dowil on plots'of land near the scene of their work, made over to them for home cultivation. In the west of the province, too, there is a flourishing colony of the same class, under missionary^ supervision. An attempt to relieve parts of Bihar by means of State grants of land in Burma has been made within the last few years, and though it is too soon as yet to judge of the results, the progress made on a small scale seems encouraging. The difference in race and language, however, will pro- bably be found an obstacle to any great extension of the movement. Temporary displacement of the population, however, both agricultural and other, is by no means rare, and is increasing in both number and variety. The movement is not yet extending to any significant degree to foreign countries. For instance, during the last 10 years, the number of emigrants registered was only 130,483, the greater portion of whom were bound for the West Indies and Guiana. But to Burma and Ceylon from Madras, and from Bengal to the tea gardens of Assam, the annual movement is con- siderable. The docks and mills of Bombay and Calcutta, too, and the wheat harvests of various parts of northern, and central India, and the plantations of Coorg, attract a considerable number of, extra hands, the proceeds of the labour serving to sustain the whole family during the period when the cultivation of their plot of land at home does not entail their presence. There is then to be considered the accretion to the population due to immigration from beyond India. This movement is mostly confined to the tracts immediatdy bordering on Nipal and the north-west frontier, and most of the immigrants come from the first-named State and from Balochistan, Afghanistan, and the strip between the last and the Panjab, called Yaghestan, or the independent territory. We have no corresponding returns from these countries to show how far the inward movement is balanced by one in the opposite direction from India, but where the boundary is merely political, and denotes no distinction of race or customs, the larger proportion of the movement is probably restricted to villages immediately contiguous to the frontier line on both sides. The marginal statement shows how small a part immigration, plays in the con- stitution of the population of India. Taking 10,000 of the people at random, only 23 will be found to have come from across the frontier. Of these last, the nine from Nipdl are chiefly along the frontier of Bengal and the north-west, and the rest in the Native Army. Balochistdn supplies only Sindh, except for a few Mekranis and others entertained as guards in native States. The natives of Afghanistan and its neighbour are mostly on the Panjdb frontier, or scattered m small bodies of pedlaifg all. over Birthplace. Number. British Provinces Feudatory States Total within India Balochistan Afghanistan Taghistan Nipil Asiatic Countries remote from India Else-where 0. 67 Table B. northem and central India. Of. the other Asiatics, one will pijobably be a Chinese, ^d the other an Arab ; the, former in Burma, the latter in Aden. There are,. of course, numbers of other birthplaces returned, but under none but the United Kingdom do the figures japproach the proportional limit above given,, of one pei^sdn in 10,000. The statement B, below, shows the main facts ascei'tained with regard. to the birthplace of the foreign population. The full return at the present census will be found in Table XI. (A. and C.) of the general series. It should be remembered that in this statement we are dealing simply with birthplace, not with nationality. Though the relative nuinbers involved are insignificant, it is worth while to compare them with those of the census of 10 years back, so the returns for Kashmir and Upper Burma are not in- cluded in the marginal table. On the last occasion, Yaghis- tan was not distinguished from its neighbour, and there were, no doubt, good reasons at that time for the population to re- frain from entry into India. The increase of people from Nipal is mostly due to immi- gration into Darjilihg' and the adjacent territory, for work on tea gardens. There are settle- ments, too. of this race in the Korth-West Province Tarai, as weiU as in Benares, the object of the aspirations of all devout Grurkhas. Passing over difiBr- ences which are probably duo to classification, we come to China, which has furnished a comparatively large contingent to Rangoon since last census. The addition to those from Arabia is due, chiefly, to the development of Aden, which accounts, too, for the great number of those born in Bast Africa. The bulk of the per- sons returning birthplaces in Europe beyond the United Kingdom, i^ to be found in the centres of commerce and mis- sions. The Scandinavians, especially, affect the seaports. Of the rest, the Germans and French are the only nations strongly represented, though most of them show a slight increase since 1881. The re- turn for America is obscured by the confusion between the States and Canada, more es- pecially in the army and rail- way schedules. Australia con- triibutes a good many more 12 Birthplace.' Numbut returned. 1881. ■' l'891. ■- ' Asia. Afghanistan and Taghistan 125,141 168,656 Belochistan 60,318 61,433 Turkestan 429 977 Thibet 2,756 1,641 Bhotan . • - 4,964 4,353 Nipal - - - . 135,166 233,553 G^ylon 2,774 5,574 Further Inflia, &e. 4,662 8,259 Arabia, &c. 23,430 28,065 " Persia, A6. Mangalore - + 27-49 17. Ajm6r + 41-26 18. Hubli + 43-40 19. Bhaunagar + 20-63 Taking first the modern or commercial towns, there is a point that must strike Modern or one as remarkable at the verj outset. This is the small increase in the two leading commercial cities of India. Whilst seaports like Karachi and Rangoon, which are centres of °^°*'®^- distribution and nothing more, have increased by 43 and 34 per cent, respectively, Calcutta and Bombay, which are not only great centres of trade, but also manufacturing towns of notable prosperity, have only an increase to show of 8J and 6|- per cent. The explanation is the same as could be given for liondon, so far as , Calcutta is eoncemed ; namely, that the tendency is now to live in the suburbs, where rent and food are cheaper, taxes less burdensome, and with the extension of railways, access to the city within the reach of all. Bombay is not so fortunate in its surroundings aS Calcutta, but the middle classes engaged in that city have betaken themselves to residences some distance out of the town in far greater numbers than were shown at the census of 1881. But the main cause of decrease, or rather of the reduction in the rate of increase, is the withd iagal of wflDien fron a^ BamMjLJn thejigcade. Whether there is less demand for their labour in the cotton and other mills, or whether the greater facilities of intercourse with the coast have induced nien to keep their wives at home during the open season, and to indulge in an occasional run to visit them, instead of bringing them up -with their children to the more expensive life in the city, is not stated in the reports hitherto received. But, at any rate, the fact remains that a considerable number of the women who were to be found in the city in 1881 were located in the Ratnagiri district at the time the last census was taken. L I 78388. 82 Provincial and military centres. State- centres. Keligions- centres. City. Per-oentage of City-born. IVr-centage of Variation. 1K81. 1891. City- Total born. Population. Bombay Calcutta Madras Eangoon 28-41 27-42 72-68 36-41 25-10 30-49 71-97 33 '68 - 4-17 + 15-67 + 10-40 + 24-29 + 6-28 + 4-10 + 11-50 +34-39 The marginal statement will serve to show clearly the difference between the four chief cities in respect to their Table N. relative attractiveness to outsiders. Madras, for instance, in spite of its large increase, is chiefly maintained by local prolificity, whilst Bombay breeds only a quarter of its popu- lation, and neither of the others more than a third. But it is only in Bombay that the home-born show a decrease. In Madras there is a tendency towards immigration, but the natural increase has nearly kept up with the total growth. In Eangoon, on the other hand, the former is remarkably high in rate, but the immigration from outside the province quite outtops it. The figures for Calcutta are open to doubt, as the areas have been much changed since 1881, and the last column clearly refers to the old city alone, irrespective of the suburban portion that has since been included. But the increase of the relative strength of the city -born is not affected by this change, and it is here that this town differs from the rest. The other towns in this class may be dealt with more generally. Cawnpore is the rising commercial centre of the fertile tracts under the government of the North-West Provinces. It is increasing, too, in manufacturing importance. Amongst other innovations it has taken the lead in the leather trade of India, and with such success that much of the work that used to be done for the army and police in England has been assigned to the manufactories here. Dehli, top, is not left stranded by the stream of modern commerce. Like Cawnpore, it is (the great wheat market of the surrounding tract, and manufacturing industry also is alive there. It has, moreover, the prestige of former metrdpolitanism, and is well known to many of the country folk all over north and central India, to whom Calcutta and Bombay are names of no significance. Calicut, Cocanada, Mangalore, and Negapatam are all seaports of rising trade. Maulmain has been outshone by its younger sister Rangoon. Bhaunagar is not only the chief port of the peninsula of Kathiawar but the seat of a wealthy feudatory Thakur, or Chief. It has thus the advantage of both commerce and patronage. Hubli and Trichinopoli resemble Ajmer in owing much of their prosperity to the extension of railways. Howrah is part of Calcutta, and Sholapur, which began well as a cotton centre, seems not to have recovered from the results of the famine, which pressed very severely on the weaving classes of the Deccan. About the second group of towns there is little to be said. Poona has the advantage of being both a political and a military centre, and has a good deal of the prestige of former times, like Dehli. R^wal-Pindi, which shows so large an increase, is purely a military station, and probably some of its rise accounts for the com- paratively small increase of Peshdwar, which has a good deal more trade but less garrison, now that Pindi has been found more advantageous from a military standpoint. /^ The third class of towns can he taken with the remarks made in^he last chapter as to the position of the capital of a Native State. Patna, Surat, Mirzdpur, and Tanjore, which have descended from that eminence, are all on the Avane. Lucknow Agra, and Dacca show less than half the general rate o|/ increase. There are no doubt, special reasons to account for the slow progress of |lampur and Jddhpui;,^ since the latter at all events might have been expected with the increase to have been more in sympathy with the surrounding State. Finally, the centres of religious pilgrimage claim a little comment. Of the six given in the table, Madura, perhaps, should hardly be taken in the same line as the rest, since it has its reputation as 'a place of trade as well as its sacred associations to support it. If we exclude it, the general ra;fce of growth of the class is ^-ery slow. In the case of Amritsar, 'there is, as has been already mentioned, a special cause for the falling off in population and it is possible that in addition to the epidemic of fever nine years or so ago, there 'may have been a loss by the return to Kashmir of some of the artizans who fled 'that State in the famine of 1878. On the other hand, as regards all towns of this sort, it may be said that no rapid increase is to be expected, since there is a temporary influx of purchasine power at periodic intervals, perfectly well known beforehand, so that the vendors o^ 83 all goods that can be easily transported mereij join the throng of visitors at the times in question, and do not waste their time in residing permanently in a place where trade is so spasmodic. The increase of the population has hitherto been discussed merely in its relation to Parallel the number of people who were enumerated 10 years ago, and this is, of course, the movement of only aspect with which the census has to deal. But there is another of equal pop"l**'<"} importance, but which, as the facts concerning it do not arise out of the operations p°osMrity * under review, can be only treated here in the barest outline. Still, it will not be altogether out of place, as it bears on the question of the probability of the continuance during the current decade of the rate of growth that has prevailed in that which has just passed. The point is, whether, judged by the facts by which the progress of a country from a material standpoint is usually tested, it can be fairly .assumed that in India the increase of the population has been accompanied by a parallel development of its resources. The simplest way of treating so complex a mass of statistics is to make use of a base-line or index-year. It may be admitted at the outset that for anything like a detailed examination of the conditions such a method is faulty, but our present object is to give a sketch only. The table given on the next page, then, shows the figures for each year of the decade in their relation to those of the year which imme^ately preceded that period. Tor instance, taking the first column, the land revenue of 1880-81 being 100"00, that for 1881-82 is represented by 103*96, and the average for the decade by 107'94. On the other hand, the imports of kerosine oil were only 90"96 in the first year of, the decade, but rose to 542-30 in the last. To save space, however, the lOO'OO has been omitted in favour of the appropjiate sign. Now, the increase of the population in British provinces is 9'70 per cent., and what with recoveries from famine and the greater accuracy of the returns, it may be safely assumed that this rate is above the normal, and the latter can be set down at about 8*50 for this territory as a whole. This, then, is the ratio with which the figures should be compared in the several columns relating to revenue, post^ and currency. In the case of the States the rate is no doubt higher, and may approach lO'OO ; and with the latter we may take the figures regarding railways and trade generally, since the whole country contributes to their support. In the former case, that of the Provinces, it appears that the increase, except in the assessment on the land which, for reasons that are too lengthy to be stated here, moves slowly, is generally above that of the population. Salt, for instance, the only tax paid by the lower classes, shows a remarkable increase in the amount consumed as compared with the revenue from it, taking the whole of the 10 years together, though the two last years of the period indicate a less satisfactory divergence, which proved, however, only temporary. The postal system shows a marvellous development, especially in the matter of cards; and telegrams, too, are now spnt to an extent unheard of even 10 years ago. In connection with the railway passengers, it must be noted that in India the trafl&c by the upper classes is comparatively insignificant, so that the bulk of the growth is to be set down to the third. It is also noticeable that the rate of increase in both passenger and goods traflBc is considerably above that in the open mileage, and the same remark applies, though less forcibly, to the net earning of the lines. "With reference to the trade returns, it is perhaps advisable to present the actual as well as th^ proportional figures, so as to indicate the general distribution of the whole amount entering and leaving the Table (i = looo). Imports. Exports of Indian Products. Total. Ex. 57,787. Total. Ex. 86,491. Cotton Goods - Cotton Twist Hardware Iron Copper Oil Silk Goods Sugar ' - Woollen Goods - Others; 23,732 3,650 921 2,110 1,837 1,470 1,393 1,866 1,403 19,405 Eaw Cotton Cotton Goods Cotton Twist - Hides Wool - Opium Tea Oil Seeds Eaw Jute Jute Goods Wheat - Kice Others 14,762 2,437 3,601 4,731 1,302 10,786 4,668 9,225 5,9SS 1,734 7,172 9,102 11,058 country. This is done in the marginal table, which represents the average values for the 10 years in question. Speaking generally, the import branch is so much niore varied than the other, that the articles mentioned, though the most im- portant, form but two-thirds of the whole value. On the other hand, the chief articles of export leave a comparatively little remainder. In the proportional tables quantities, it should be men- tioned, are. taken instead of values, owing to the complications introduced into the latter by the vagaries of exchange of late years. On the whole, it will be noticed that as to merchandise the increase in L 2 84 Table P. Variation pee Cent, from ^he Returns for 1880-81 of Main Items of Revenue, Commerce, &c. State Revenue from 1 ' Post and Telegraph Transactions (Non-Offloial Correspondence only). 1 1 Bailways. Private Trade ■( ralifey. ' gig II 1 Merchandise. Treasure. la. ik Tear. 1 ^ *3 1 ■s 1 D 1 i 1 i| 1^" 1 5 "S t 1 s 1 S Exports (of Indian Produce and Mariufac- tiires). i p. a 1881-82 + 3'96 + 4-02 + 3-52 + 3-49 + 5-97 + 47-68 - 0-86 - 1-16 +. 4-16 + 12-61 + 12-2S + 9-80 - 6-S9 +10-12 + 2S-97 -23-65 +66-26 1882-83 + 3'6l + 3-97 -13-18 + 8-52 + 17-63 +100-77 + 13-92 + 11-11 + 7-80 + 24-51 + 29-36 +21-21 — ' o'6i +11-98 + 4Q-68 -27-66 +28-25 1883-84 + 5'92 + 8'08 -13-63 +10-32 + 28-16 +169-92 + 18-39 - 2-02 + J2-S4 + 41-18 + 26-13 +J3-77 + 4" 76 +18-22 + 43"2S -29-88 +36-53 1884r-85 + 3'4l' +10-95 - 8-56 +13-66 + 3S-87 +227-60 + 30-52 + 6-42 + 17-00 + 64-67 +43 -25 +31-26 + S-6S +11-59 + 54-52 +36-81 + 0-78 1885-86 + 7'oj +12-69 -10-83 + 12-03 + 53'20 +295-92 + 46-34 + 7-67 +28-07 + 69-15 +48-18 +40-57 + 2-99 +12-17 + 72-20 -23-06 -13-J8 1886-87 +. 9'20 +15-40 - 6-44 +19-05 + 63-60 + 355-81 + 61-26 + 3-94 +42-72 + 82-49 + S2-87 +34-69 + 16-60 +17-97 + 22-9S +19-44 + 6-40 1887-88 + 9-83 +19-25 - 6-26 +16-06 + 72-94 +410-29 + }8-i8 +18-29 +46-11 + 97-30 +69-J0 +42-25 + 24-00 + 20-00 + 53-82 +11-40 -24-8S 1888-89 + 9 "02 + 20-81 + 5-86 +17-00 + 84-80 + 477-96 + 102-6J +20-26 +46-76 +111-64 +68-41 +45-50 +32-32 +28-72 + S4-03 +23-88 -16-83 1889-90 + 13'S9 +25-76 + 1S-06 +16-98 + 96-63 + 535-30 + 109-90 +18-21 +70-64 +118-20 + 71-16 + 49-02 +32-30 + 37-69 + 94-2S +32-44. - 3 "79 1890-91 + 13-89 +25-18 + 19-78 + 18-29 + J0S-01 + 579-80 + 126-51 ■ +67-53 +79-16 +133-22 + 96-77 + 82-50 +37-22 +33-25 + 144-04 +47-40 -64-25 Mean of) Ten >• Tears J + 7"94 + 14-61 - 2-23 + 13-54 + S6-31 + 319-11 + S8-70 +15-02 +34-86 + 74-50 + Si'79 + 37-06 + 14-86 +20-17 + 61 '47 + 6-70 -1-83 Imports (Quantities). Exports (Quantities), Indian Produce or Manufacture. Year. Iron. Hard- ware. Copper. Keroslne Oil. Sugars. Cotton Piece Goods. Cotton Twist and Tarn. Raw Cotton. Rice. Wheat. Oil Seed. Raw Jute. Raw Wool. Cotton Twist and Tarn. Cotton Piece Gtoods. Jute Manu- factures, 1881-82 1882-83 1883-84 1884^85 1885-86 1886-87 1887-88 1888-89 1889-90 1890-91 - 7-99 + 18-25 +32-94 +35-14 +31I0S + 23-06 +62-12 + S0-J7 +35-37 +45'43 + 13-40 + 48-30 + 47-29 + 62-84 + 40-64 + 56-62 + 97-98 + 99-*7 + 98-39 + 116-74 -11-42 + 17-92 +38-92, +44''73 + 71-08 +61 - 14 +39"5S -74-20 +49'o5 + 16-97 - 9-04 +105-36 + 35-23 + 171-34 + 81-64 +219-47 i +211-59 +295-01 +434-85 +442-30 - 21-33 - 31-80 - 25-29 + 63-93 +.18-74 + 77'38 + 83-36 +■ 64-01 + 74-70 + 197-26 - 8-56 - 7-63 - 2-96 - 2-,S9 - 1-86 +21-35 + 3-82 +19-70 +12-42 +13-39 -11-15 - 2-22 - 1-09 - 2-35 + o-oS + 6-83 + 12-35 + 14-63 + i-io + 11-10 +23-9S +35-86 +31-83 + 11-63 - 7-71' + 19-69 + 18-3S + 17-40 +39-23 +30-46 + S-96 +14-64 - ,0-83 -19-12 + 3-61 - 1-42 + 4'66 -16-12 - 0-61 +28-23 + 167-33 + 90-66 + 182- II + 113-92 + 183-02 + 199-07 + 81-86 + 136-57 + 85-40 + 93-37 + 1-73 +27-60 + 68-46 +77-22 + 68-09 + 64-38 +56-08 +61-13 +63-32 + 43-66 + 29-27 + 78-13 + ,20-80 + 44-04 + 33 -95 + 42-98 + 65-99 + 81-64 + 76-53 + 105-30 + 3-92 + 2-46 - 1-99 - 0-86 +21-67 +31-07 + 36-26 +36-39 +48-64 + 32-67 + 14-44 + 68-68 + ,85''4i + 144-96 + 190-85 + 241-26 + 331-73 + 379-I8 +427"67 + 539-34 - 1-63 + 36-69 + 82-87 + 57-67 + 69-68 + 75-62 + 128-62 +131-18 + 95-70 + 122-61 -19-69 +27-40 + 31-49 +58-02 +21-71 +33-36 +41-96 +gor49 +85-90 +88-50 Mean of) Ten Tears J + 32-56 +66-66 + 26-68 +198-77 + 50-10 + 4-71 + 2-93 +32-07 + 1-99 + 133-13 +60-16 + 57 ''96 +21-01 + 240-34 + 79-88 +43*9' 85- the exports has been continuous, and in rate above that of the imports, whilst, including treasure transaction, the balance is inclining a little in favour of closer correspondence. But here, as in the case of the post, telegraph, and railway, one of the main points worthy of note is the increase in the apparent reserve available for spending on what are to the masses of India, objects of luxury. Then again there is to be noted the growth of the export trade in raw produce, though it is here as every- where else, subject to annual variations of a somewhat irregular character. Lastly, the competition between imported piece-goods and yarn and that made in India. As to the first, the marginal table shows that the absolute quantities manufactured abroad and sent to India are very largely in excess of the others, but the proportional statement indicates the relatively far higher increase that has taken place in the exports of this sort of goods frOin India in the 10 years under review. The yarn from abroad, again, is being beaten in the race, just now, by that made in the country, but there is probably a check to be expected in this rapid increase of foreign demand, as there has been to a slight extent in the jute trade of late. Finally, a few words must be said about the absorptive power of India for the precious metals, which, in discussing the circumstances of 'the country, is a point too often overlooked. In the last 10 years India took in an amount of gold which is valued in what is known as Bx, or "tens of rupees," at no less than 44,051,255, whilst it disgorged in the same time only 3,144,069. Simultaneously, the corresponding amount of silver introduced intb the country was 101,086,766, of which only 12,225,899 has left it. Thus there remains in round numbers, the equivalent of Ex. 41,000,000 in gold and 89,000,000 in silver, which has been added to the hoards already in the possession of the people previous to the period under review. We have every reason accordingly to assume that the present rate of increase amongst the people of India is well within their means of subsistence. If maintained, which of course it will not be, it would be 75 years before the population doubled itself, and the problem of their support would then, no doubt, be a hard one for our successors. L 3 86 Special diffi- culties met witli in enu- merating occupation. Scope of the inquiry. CHAPTER IV. The Occupation of the Population. " By ceaseless action all that is subsists. Constant rotation of the unwearied wheel That nature rides upon, maintains her health, Her beauty, her fertility. She dreads An instant's pause, and lives but while she moves."- -Cowper. Tlie Statistics that have been dealt with, in the foregoing chapters of this review are of a comparatively simple character. In some particulars inaccuracy is admittedly present in the returns, but it is not diflBcult to trace, and it can be localised within certain fairly well defined limits, so that its effects can be adequately discounted in discussing the points in question. Uncertainty, therefore, as to the true meaning of the statistics prevails to but an insignificant extent. For ^example, the deliberate omission of ■ adult females is attributable to a single cause operative amongst certain classes or within certain tracts. The return of birth-place, again, is well known to be often characterised by excess of minuteness on the part of the native of India who happens to be sojourning at a distance from his home, and by equal want of definition on that of the foreigner. The one insists on entering his native village, without adding the name of the State or district in which it is situated, whilst the other too often considers that the name of a continent is sufficient to satisfy reasonable curiosity on this matter. A still more striking case is that of the age-table, with which we have yet to deal. Here, the prevailing ignorance is so uniform and so widely spread, that the error tends to group itself into certain easily recognisable forms throughout the whole country, so that in accordance with the general tendency of such large aggregates, the very irregularity of the figures becomes, to a considerable extent, the test of their accuracy. But with regard to occupation, the conditions are more complicated and the opportunities of going wrong proportionately increased. To begin with, the diflficulty in formulating the question is considerable, for the circumstances are so various that it is hopeless to attempt to frame instructions or illustrations comprehensive enough to cover the whole subject. Then, too, there seems to be in all countries alike a tendency towards certain classes of error, generally in the direction of the use of terms too indefinite to indicate sufficiently the precise nature of the occupation, or which bear a special local signification alien to the usual meaning. Lastly, owing partly to the difiiculty of the subject itself, partly to the defects above mentioned in the original record, a return of occupation may be, and too often is, grievously handicapped by the system adopted for the grouping and classification of the enormous arabunt of detail that has to be thus worked into shape. As regards the first, difficulty, the main point is to settle exactly what is to be the scope and extent of the inquiry and this necessarily differs in every country, according to the stage of its development. In all cases it is held necessary to ascertain the numerical proportion of the main classes of occupation, such as agriculture, manufacture, mining, commerce, or the liberal professions, not to mention the less definite ca,tegory of the unskilled, which has sprung into such prominence in some countries within the last few generations. In most instances, too, it is advisable to estimate the relative strength, in the majority of the above classes, of possession and proletarianism, the wage-paying and the wage-receiving popiilation. But the question then arises whether the return is to be that of the working members of the. community only, or that of the supporting power of each means of subsistence at the time of the census. In the first case, it would include only those who actually work, or who are in the immediate receipt of the means of livelihood which render them independent of work. In the other, the scope is widened so as to comprise not only the workino- members of the househoM, but all who immediately depend upon them for their support. Similarly, in the case of those who have the means of living without the need of working for them, the return would include, in addition to the person actually in possession of the income, those supported out of it by him, otherwise, of course than by way of charity or wage for special service rendered, since the last named are not indissolubly connected with the income in question, but constitute independent means of livelihood. The relative advantages of these two rest mainly on local considerations. The reason for selecting the latter, or wider, system for India will 87 be found later on in this chapter. Another question under this head is the amount of detail in returning the occupation that should be asked for. Whether, that is, a general title or class is considered enough, or whether greater accuracy should be demanded. The answer to this depends, obviously, on whether the object is to get a complete industrial survey, or merely a general view of the distribution of occupation on broad lines. Connected with this point, too, as we are on the topic of difl&culties, is the much vexed question of how to deal at the census with the numerous instances in which a single individual or family, as the case may be, has more than one means of livelihood, or where, as frequently happens, and not in India alone, the return is that of a title that may include all or any of several occupations. The second of the above mentioned obstacles to a complete return will be found inaccuracy- well illustrated in the course of examining the details of the figures for India, though, in the entry as was hinted before, the inaccuracy in question is by no means confined to that °f occupa- couatry. A few entries o£ this sort, however, may be here quoted in explanation of the '°°' difficulty. "We have thus a lot of people returned as " shopkeeper," without any guide as to the goods sold. " Labour," again, leaving it uncertain whether it is agricultural that is meant, or the general unskilled labour that chiefly haunts towns and extensive public works. " Service," is another term that is used without qualification, more often, probably, in India than elsewhere. It may there be that of the State, or of the household, or of the shop, or in a merchant's establishment. The profession of a "writer" is as common in India and as various in its application as that of " mechanic " in England. It has no connection, however, with authorship. Numbers of others of the same description suggest themselves, but it is superfluous to quote them. The general question of the classification of occupation is a highly complicated Objects of one, so it is not proposed to enter into it minutely in the present work. The classification. distribution of industry varies from country to .country so much that comparison is both valuable and interesting, but the field of inquiry is so extensive and full of detail, that however desirable it may be to make such a comparison, for which, of course, a general scheme of classification is essential, it has been found impracticable hitherto to carry it out in a trustworthy manner on the basis of a census return alone. The subject, therefore, has not been recognised as one falling into the category of international statistics, and there seems abundant reason for not placing India side by side with western countries in such a return. If once we abandon the idea of international comparison, the object of classification is confined to the exhibition of the leading ■ characteristics of the local population with reference to their occupation, irrespective of the sociological considerations of a general nature to which weight has to be attached when the field of comparison is enlarged. How many of these have been dropped out of sight in dealing with the returns for India on this occasion will be seen from the comments made below on the system of arrangement herein adopted, but, so far as the main purpose, that of indicating the general features of the social organisation of the country, will admit, the obvious advantages of theoretical consistency have been retained. It may be gathered from these remarks that a high value is not attached to the General results of the census of occupation. This is true, and the opinion is not confined to value of the those who have had the administration of the operation in India alone. In some of the returns. . countries of Europe the subject is excluded altogether from'the enumeration, and in one, at least, which need not be named, much forethought and many elaborate instructions were rewarded by results which the census authorities thought it advisable not to mislead the public by including with the rest. In Germany as well as in the United States, it has been decided that a comprehensive industrial survey, obtained by dint of detailed inquiry spread over a considerable time, is preferable to the rough and ready return which is all that it falls within the capacity of a synchronous (jensus to furnish. The elaborate and in many respects admirable tables - that contain the results of the occupation return at the last census of England and Wales have been shown by experts in industrial economics to be incomplete in several important particulars. Italy, on the other hand, is a staunch supporter of the census of occupation, and the opinion of a statist of the eminence of Oommendatore Bodio, who has brought his country's statistics to such a, pitch of efficiency, carries great weight. Still, he would be, no doubt, the first to admit ;t^t the section in question of his Confronti Internazionali can only be digestedwith a- copious sprinkling of explanatory salt. This usuallytakes the shape of the enumeration of the various ways of going wrong, as above, which has L 4 ' S8^ the further advantage that many minds of a complacent order take the admission of statistical error as incense offered in some way or other to what they speak of as common sense. India, it is ' true, is, in its present stage, at an advantage in respect to the enumeration of occupation, as compared with countries which are more adapted' by situation or natural products to commercial or manufacturing development, since, in the former, the specialisation of functions is, relatively speaking, in its infancy ; the organisation of labour is of the simplest, and the multiplication of the various means of subsistence proceeds at a rate barely perceptible, except at considerable intervals of time. But, from a census point of view, the very simplicity of this almost exclusively agrestic community has its drawbacks," and the complication that arises^ is larg'ely due, not to the manifold differentiation of occupations, as it might be in England, but to their combination ; as, for instance, when one individual lives by the exercise of several which are otherwise totally unconnected with each other, and he is probably unable to specify which of them may be by preference termed his chief means of support. In fact, in the most favourable circumstances, the completeness and accuracy of the enumeration is a matter of very considerable uncertainty, and to those who have acquired knowledge of the subject by experience, it is scarcely possible to conceive a state of society at all above that of primitive savagery in which it would be otherwise. System of Before discussing the results of the enumeration of India by occupation, as enumeration, qualified by the above general admissions of imperfection, some explanation is necessary of the scope of the returns and of the system on which they were prepared, because important innovations have been introduced into both since the preceding census. It has been stated above that the object of the present census was to obtain a view of the supporting power, at the time of the. census, of each means of livelihood, whereas in 1881 the return was that of workers only, and persons of, so to speak, independent means. Thus no less than 53 per cent, of the population, or, as the return was by sexes, 37^ per cent, of the males, and 69-| of the females, was excluded. If, therefore, the total strength of any particular group was in question, such as, for instance, the agricultural or the artisan class, recourse to approximation was necessary, and this had to be based on estimates of the probable number of heads of families^ and of the numerical strength of their respective households. As there was no general tabulation of occupation by age, except in a few of the larger towns, the above computations had to be purely conjectural, and the postulates and methods of calculation differed in each province. But even as a record of the working population only, the results, in the opinion of the Superintendents of the census operations, in their respective provinces, were very defective, and so far as the return of working females is concerned, the figures were found by Sir W. Chichele Plowden, the then. Census Commissioner for India, to be unworthy of examination. A few facts in confirmation of this view may as well be cited, which will show that the abandonment of the forrdei' system was neither premature nor based on inadequate grounds. First of all, of course, stands the inconsistency of practice in regard to the entry of the occupation of women, which is based to a great extent on social considerations. Where there is a strong and well defined line drawn between the upper and the lower grades of society, ov,iiig to distinctions of caste or race, the women who actually do work are generally returned as workers in the lower section only, although they may be similarly engaged in various occupations in many of the higher classes. On the other hand, there is noted a general tendency to return the women of the middle classes, all over India, as following the occupation of the head of their household, whether they actually do so or not. This last fact was one of the inducements to adopt the scheme of return indicated by this inclination. Then, again, the distinction between the principal and his dependents or sharers is much valued in certain occupations, so the chief worker especially the artisan, is disposed to ignore, in making the return, the aid rendered by the others with whom he may be connected. There is a similar inconsistency with regard to occupations dependent on the land. In some cases, the nominal occupant will alone be recognised, and the rest of the family returned as labourers, whereas in others, where the participation of the soil is based on different principles, the number of occupants or shareris will include all who have any interest therein, irrespective of the position of the patriarch. The young children of agriculturists, too, are as often as not entered as graziers, though their attention is gratuitously devoted to the family live-stock only. Elsewhere, they will be found to have been omitted from the workers altogether, or returned, in some instances, as farm servants. Other cases of a similar tendency can be adduced, but, on the whole, they reduce themselves to the 89 above categories of inconsistency of treatment of subordinate or subsidiary interest in the family means of liveliliood. Now, under the present scheme of enumeration, though the number of workers is ignored, the whole population is accounted for under some head or other, with the exception of about \^ millions, who either did not fill up the column at all or returned some term which was unintelligible, or, as is more frequent in India than in the West, frankly disreputable. The object of the return renders the distinction of sex immaterial, since man, woman, and child are alike supported by the occupation in question. From the provincial returns, however, it is possible to get the number of adult males under each head, since in these the detail of both age and sex has been retained. In the Imperial series of tables it has been omitted, first, because for a considerable portion of the population in the Feudatory States and for some in Southerr India the age was not tabulated for occupations ; secondly, because, as will appear in a later chapter of this review, the ages, in the crude state in which they are extracted from the schedules, are by no means correct in two, at least, of the periods selected for the occupation table ; thirdly, because the detail of age and sex, added to that of rural and urban population for British and Feudatory territory, would swell the return to the dimension of a considerable volume with incommensurate advantages ; Lastly, because, as shown above, these details are irrelevant to the general purport of the statistics in question. But although the distinction between worker and dependent is obliterated in the final results of the census, it was retained, as will be seen from the instructions quoted below, throughout the enumeration. This, however, was done merely as a measure of administrative convenience, not with the view of tabula- tion. There are numerous prejudices on such matters amongst the people in India which deserve consideration, and by prescribing the addition in this column qf the schedule of the word " dependent " to the entry of occupation made against such as do no work, the awkardness was avoided of appearing to attribute to these persons the exercise of various functions which are impossible or prohibited to those of their age or sex. Had the distinction been carried into the tables, it is very plain that the defect that vitiated to such an extent the results in 1881 would have been perpetuated, and casual readers would have been liable to be misled by assuming the number of the active members of the population to be that set forth as such. The instructions for filling up the occupation column were the fullest and most Instructions elaborate in the code, for it was decided that as this was the most difficult portion of the *» enumerft- task imposed on the enumerating agency, it was only fair to carry the explanations ^'°'^^' of what was wanted as far as the time available for their instruction and the limits of their intelligence would allow;. It was also held desirable to .obtain in the first stage of the operations as much detail as possible, so that any reduction that might seem advisable could be postponed till the whole field of the subject came under the eye of the Provincial Census Superintendent during compilation. If general terms, or classes of occupations, had been considered sufficient, the instructions could easily have been curtailed into a few lines. In order to show precisely the nature and scope of the inquiry under this head, and the difference between it and the corresponding procedure at the census of 1881, the heading of the column in the schedule and the rules for filling in the entries to be made on the two occasions are here, reproduced : — 1891. Occupation or Means of Subsistence. ' Instructions : — Enter here the exact occupation or means of livelihood of all males and females who do work or live on private property such as house rent, pension, &e. In the case of children and women who do no work, enter the occupation of the head of their family, or of the person who supports them, adding the word " dependent," but do not leave this column unfilled for any one, even an infant. If a person have two or more occupations, enter only the chief one, except when a person owns or cultivates land in addition to another occupation, when both should be entered. / 78388- M 1881. Occupation of Men, also of Boys and Females who may do work. N.B. — Boys at School, Girls, Small Children and Women who perform no regular Work, should not be ihown at all in this Column. Instructions : — Only, such persons are to be shown in this column as actually do work contributing to the family income. Mere employment in such do- mestic occupations as spinning will not entitle women to be shown in this column, unless the produce of their labour is regularly brought to market. When a person has two or more occupations, he should be entered as following the occupation whence his income is chiefly derived, 'but if he combines agriculture with any other profession or trade, such as that of vakil or money-lender, carpenter, or smith, both occupations' should be shown. 90 1891. (continued.) No vague terms should be used, such as service, Governmsnt sei-vifie, shopkeeping, writing and labour, &e., but the exact service, the goods sold, the class oi' writing or of labour must be stated. When a person's occupation is connected with agriculture it should be stated whether the land is cultivated in person or let to tenants; if he be an agriqultural labourer, it should be stated whether lie be engaged by the month or year, or is a daily field labourer. Women who earn money by occupations independent of their husbands, such as spinning, selling firewood, cow-dung cakes, grass, or by rice-pounding, weaving, or doing house-work for wages ' should be shown under those occupations. If a person makes the articles he sells he should be entered as " maker and seller " of them. If a person lives on alms, it should be stated whether he is a religious mendicant or an ordinary beggar. When a person is in Govern- ment, railway, or municipal service, the special service should be entered iirst, and the word Govern- ment, I'ailway, or municipal, &c., after it as : — clerk. Government ; sweeper, municipal ; labourer, rail- way. If a person be temporarily o\it of employ, enter the last or ordinary occupation. 1881. (continued.') General terms, such as seryant, workman, dealer, must ijot be eniiployed. In each case the , sjpeQifiq service or trade in which the person is engaged musi be named, e.g., w'atchman,' officeimessengerj 'digger, ploughman, - cloth-seller. General expresBions [such as pesha-iTkhud'] musiti npt be employed. In every case the occupation must be indicated by;the common vernacular term by which it is known [and nbf by the Persian name ; thus Kumhdr for potter, not Kasgar']. The qhanges made in 1891 are chiefly in detail, always excepting that shown in the beginning of the rules. One or two points were treated at greater length, where it had been fouufl that the first code led to misunderstanding. For instance, the phrase " regularly brought to market" is said to have excluded a good number of the class to whom it was intended to apply. Again, in some few but important cages, the name of the caste iS' a better indication than that of the principal occupa- tion, so the penultimate provision of the older code 'was omitted. The greater detail as to the connexion with the land was intended to serve as a guide to the relative strength of the cultivating class and the non-agricultural proprietor or occupant, and a special provision was held necessary in order to prevent the entry of •" household work" against those who were not getting their living by it as an independent source of livelihood. The bulk of the rest of the modifications simply met cases of doubt that had arisen in the course of the operations of 1881. It should be mentionedj by the way, that the new rules were framed for the sanction of the Government of India by a Committee of the census Superintendents of 1881, appointed for the purpose on the suggestion of Sir W. Chichele Plowden, the Census Commissioner on that occasion, as affording the best means of making use of the experience then o-ained; There' are two points, besides that of the obliteration of the distinction of the dependent classes, which are of sufficient importance to justify a few explanatory words in this place, as they appertain to matters of somewhat wider interest than those above-mentioned. The first is the application of the final sentence of the 1891 rules to the inmates of prisons or asylums. The general instructions, as they stand, seem to imply that at all events temporary prisoners and inmates of asylums should be returned according to the means whereby they lived before their withdrawal from public life. On the other hand, the special rule issued to those in charge of the census of the above institutions prescribed that the inmates should be returned as such, irrespective of their former or ordinary employment. The discrepancy is based on two arguments. First, that the results of the census represent, so to speak, the people photographed at a given time without reference to that which preceded or was to follow it. Clearly, therefore, the persons in question were not at that time exercising lany occupation by which they lived, but were, for the moment withdrawn from the active section of the population. Then, again, we should look at the source of their means of support, which ia the taxation of the public, or private charity, as the case may be, but not the result of their own labour. They are thus in an entirely diff'erent position to that of the free agent temporarily out of employ, who is presumably living on the proceeds of previous work of the sort he intends to resume. A somewhat kindred point which has occasionally arisen is that of the entry of the original occupation by those who have ceased to exercise it. Here, too, there is a withdrawal from work, and the means of 91 subsistence are either investments or:j at best, capital derived ttrough. a past occupation, just as; witk reference' to the present question, the officer of> the State is in a different position when he is in receipt of salary from that he occupies when on pension, quite apart from 'th6 celebrated definition of the latter in Dr. Johnson's Dictionary.* The seeond matter which seems to be open to comment is the way of dealing wi'fch, the people who have more occupations thaa. one. It will be seen from the rules that only one class of such cases has been recognised^ namely, where the second means of livelihood is connected with the land. This is in India by far the most prevalent form of such pluralism ; the next to it is, perhaps, money-lending, which is practised:, ta a greater or less extent, by. most m^n of , substance, 'pxcept ]\^usalman, and tlien comes a misoellaneous group of occupations wHoh are- associated under the name of Jhe^caste which exercises them. Both these last have been lost to us in the census retiurn, owing,: to the selection of, the main employment only, and as regards the first, of the twOi it is very doubful whether the return would have been anything like pomplete had it been prescribed, because the inquisitiveness about subsidiary sources of income w;ould possibly have raised suspicions connected with taxation, from which, as it was, the proceedings were fortunately free. The omission of the seeond is undoubtedly to be deplored. As to the general question, it must be admitted that the rule followed in Biigland, of entering all the occupations in their order of importance to the. total income of the family or householder, is the best, though for the reason above given, it may seem inadvisable to apply it at present to India. Such was, in brief, the scope of the information it was sought to obtain. After Abs>traetion the census, the abstraction from the schedules was effected for the rural and' urban «* ^^'^ tracts respectively by sexes, showing for e&ch the distribution of every occupation by '■®™^*^- three age periods ; first, those under five, which would include all who are too young to work ; secondly, between five and fifteen, an interval which covers the school-going age of those of the upper classes, and that of casual help in the home occupation on the part of the youth of the lower and middle. The third, which includes all over fifteen, though it contains probably a good many of both sexes who are either still 9,t school or who are not engaged in regular work, represents, as fairly as can be expected, the working age of the masses. A subsidiary object in selecting these periods should not] be forgotten, and this was the check afforded by this independent tabulation on the age tables. It has been stated above that in the Imperial tables no use has been made of the details of sex and age, but in most of the. Provincial reports comments have been made on them, by the census Superintendents. It requires note, too, in connection with the present topic, that under the general instructions for abstraction, the whole of the detail found in the schedule was recordedi, so that all grouping could be handled by more experienced agency at a later stage in the operations. , laists of all doubtful cases were sent for decision to the Census Commissioner, and in several Provinces the complete index was so treated. The object of the above provision was obviously to ensure as much uniformity Classifica- in the final return as the nature of the information allowed. To further this end, tion. the scheme of classification, which is explained below, was accompanied, by a lengthy list, showing, under their correct classification all the occupations and means of livelihood that were either returned at the census of 1881 or had been noted down in the course of tours of inspection from the schedules of that under review. They were also indexed alphabetically, in pursuance of the system adopted in such cases in the English census. The grouping and distribution by class are inevitably elaborate, but it is satisfactory to find that the scheme was found workable and convenient by those who had tte duty of applying it t'o their results. It is undoubtedly nearer to what was wanted for India than the scheme on which the returns of the preceding census was based, but it is admittedly far from being completely adapted, to its object, and the experience gathered from its use shows plainly that so far as the return of caste for occupation is concerned, and this is one of the main difficulties to be. met, the practice in one part of India is so different from that in another, that it is ,oui of the question to' deal with the matter under a single set of rules. In the Pan jab, for example, the Superintendent fotind from the returns that the number of caste entries in the occupation column were so few compared to that which returned one or other of the actual means of subsistence, that no difficulty was experienced in classifying them. On * " An allowainee made to anyone without an equivalent, In England it is generally understood to mean pay given to a State hireling for treason to his country." ' ' '- ri' J. \ • M 2 92- Classes. the other hand, the Bengal Superintendent writes of Orissa that about 20 per cent, of the population seem to have returned their caste instead of some form of agriculture, by which they really live. Thus, instead of the almost purely agricultural character of the people of this division of his charge being shown in the returns, Orissa is the only part of the province in which more than half the population are entered as non-agricultural. He points his comments by comparing the general results in this respect of the occupa- tion census amongst a Musalman population, where caste is not functional to any preponderating extent, with those of a purely Brahmanic population, such as is found in Orissa. The best and shortest way of explaining the system of classification seems to be to reproduce in extenso the instructions ' regarding it. These, then, ran as follows : — " The object in view is to group the entries in the census schedules as far as possible in accordance with the distribution of occupations in India in general, and at the same time to allow room for the designation of special features found only in certain provinces. It is superfluous, therefore, to discuss the classification in use at the census of communities further advanced in economic differentiation, or one based simply on the abstract laws of sociological science, The classification now published is not altogether scientifically correct, but it will serve its purpose if it collects under one head occupations known to be a.kin to each other, and keeps apart others which are but nominally related. " 2. There are certain classes of occupations in India which require a few general remarks before the details of the scheme are reviewed. In the first place, Government service is so comprehensive a term in this country, that for the purposes of classification it is necessary to restrict its application to the functions which cannot be dissociated from the main end of administration — protection and defence. Thus, special functions undertaken by the State in India beyond the primary duties above quoted are to be classed, not under the head of Government service, but under their special designation. Public instruction will come under educa- tion, and engineering, meteorology, agricultural training, medical practice, and administration under these heads respectively. It will be almost impracticable to affect a complete separation from the general title to which objection is raised above, as the combination of these special functions with that of the military or civil service of the Crown has been retained too closely in the schedules to admit of discrimination ; but, as far as possible, the principle above enunciated should be rigorously applied. It is the same with the service of local and municipal bodies, where only persons actually engaged in administration should be entered under those titles. Engineers and road overseers or supervisors, sanitary inspectors or surveyors, schoolmasters and vaccinators, all have their special groups, irrespective of the source from which their salary is drawn. If the extent to which in India the functions of the State are exercised beyond the limits of protection be in question, the better source of information will be the periodical lists published by Government of its employes rather than a census return. " 3. A second class of occupations needing special treatment is the very large one of what have been called "Village Industries," one great characteristic of which is that the same person both makes and sells. Amongst the most important of these come the brass-rsmith, blacksmith, cotton wfeaver, potter, tanner carpenter, and the like, representing, with their fellows, the bulk of the artisan class throughout the country. Owing to the extension of towns, it is misleading to group such occupations under what would be othersvise an obviously suitable title, and some artisans indeed may have totally changed the character of the occupation on emigrating from the simple community to which they originally ministered. It has, therefore, been thought advisable to make no difference in the classification between those who make and those who sell special goods though in the sub-divisional groups there is room for the general dealer, the commercial ao'ent, and other middlemen, and also for that class of dealers known by a special name in each province, which supplies certain articles which are almost invariably associated together throughout the country. " 4. After the above general remarks, the scheme may be taken up in detail. In the first place the aggregate of the various means of livelihood are divided into the following main classes : — " A. — Government. " B. — ^Pasture and agriculture. " C— Personal services. " D. — The preparation and supply of material substances. " E. — Commerce and the transport of persons, goods, and messages, and the storage of goods. " P. — Professions, learned, artistic, and minor. " G.— Indefinite occupations, and means of subsistence independent of occupation. "Of these, the first and fourth are the most complicated, though, making allowances for the defective return in certain cases, the former should be nearly freed from all but those who can rightly be classed in it The fourth has had to be minutely sub-divided lest confusion should arise. "5. Subordinate to the seven classes come 24 Orders, as shown marginally, bracketted according to their respective main heads. The first few Oebees. explain tliemselves. As regards the seventh, eighth, ninth, tenth, and eleventh, the object the article or service is intended for is placed more promi- nently than the material dealt with. On the other hand, from the twelfth to the seventeenth, the latter is regarded as more characteristic of the occupation than the object for which the prepared article is intended. The distinction is, of course, conventional only and not economic, as in both Orders the makers and the sellers of an article are ccm- mcution IS made of those who return themselves as A. C. D. T. Administration. II. Betenee. III. Foreign and feudatory State service. IV. Cattle-breeding, &c. * V. AKriculture. VI. Personal services. VII. Pood and drink. VIII. Light, tiring, and forage. IX. Buildings. X. Vehicles and vessels. XI. Supplementary require- ments. XII. Textile fabrics and dress. XIII. Metals and precious stones. XIV. Glass, pottery and stone ware. XV. Wood, cane, and leaves. XVI. Druffs, gums, &c. XVII. Leatlier. jj (XVIII. Commerce. ■ I XIX. Transport and storage. XX. Learned and artistic pro. fessions. _ XXI. Sports and amusemeius. f XXII. Complex occiipalinns.* G. < XXIII. Indefinite occupations. I XXIV. Independent of work. * In the Imperial Returns this has been omitted, and Order XXIII. divided into XXII. Geneiul labour, and XXIII. Inde- finite or disreputable. bined, and it is only in the eighteenth that special exclusively engaged in distribution. 93 " 6. The classification next passes into sub-orders, and where still further definition is thought necessary, Sub-orders and into groups below the sub-orders. Of the latter there are 77, which are shown, with their groups, in groupg. Appendix A. In some respects they are the most important items of the scheme, and it is possible that with careful classification, the Imperial tables may be based on them, leaving detail below groups for supplementary or Provincial returns. At all events, beyond a few generally prevalent occupations, it is probable that each Province will be best served by being given discretion to select under each group the items it considers most typical or otherwise important in the constitution of its population. Before this is done, however, every occupation and means of livelihood returned will have to be catalogued for classification, so that uniformity up to the point mentioned above may be ensured. " 7. In Appendix B. is given a sample of the application of the scheme to a collection of items found in Oceupationi. two or three of the Census Reports of 1881. It is not to be confounded with a complete index, such as was attempted on the last occasion, but will nevertheless serve as a general guide in preparing the detailed catalogues prescribed for the present census in Appendix B., page 12 of Circular M. " 8. A very important point to be dealt with in connection with the tabulation of occupations in India is Combined how to deal with an occupation shown conjointly with some description of agriculture or with the possession occupations, of land, or, again, which is locally known to be always combined with another non-agricultural occupation, of apparently quite a distinct character. It seems advisable to show all of the former class in the main return under their respective special headings, and in a supplementary return to show them with the agricultural connection as the main head, and the special occupation subordinate to each several sub-division of the former. For example : — We may have a pleader who is a non-cultivating landowner, a money-lender who is also a non-cultivating landowner, a carpenter who is a cuUivating tenant, and so on. In the general return these will appear under the items of pleader, money-lender, and carpenter respectively. Jn the supplementary return, after the total number of non-cultivating landowners who have no other occupatiohlfeturned against them, will come the pleader, and after him the money-lender, whilst the carpenter will occupy a corresponding position under the head of cultivating tenant. In this way the total number of landowners and tenants and their families, so far as the census return is con-ect, will be, obtained without detracting from the roll of the occupations which probably take up an equal or greater portion of the time of the person returning both. As regards the second-class, no general heads can be prescribed, as the combinations may differ in every province. Notorious instances are those of the tanner and shoemaker, shepherd and blanket- weaver, and, as shown by Mr. Ibbetson for the Panjab, the fisherman, water carrier, and public cook. Where the caste is used to denote the occupation, there need be no difficulty in providing a special heading for the complex functions in question. In other cases, local knowledge should be called in to point out which are the occupations almost always combined together, and these can be demarcated by a special note to the return. The instruction, -however, that only the main occupation should be entered in the schedule, is against the chance of obtaining a complete return of non-agricultural combinations. All the same, provincial Superin- tendents should suggest as soon as possible the heads of this class which they find can be distinguished in their respective Provinces." It ■will now be well to indicate in what particulars the above prescriptions failed Defects in in their object. This can be done in very general terms. In the first place, the enumera- vague title of " service " was not entirely driven out of the field, in spite of all the *^°"" precautions taken against it, so that the groups of State employ and domestic work must be somewhat intermingled, to the detriment, possibly, of the commercial groups. Then, again, the provision of food is understated, since many of those purveying grain and pulse are to be found amongst the shopkeepers, in the trading class, because their entry was of a general nature, wanting the specification of the article in which they dealt. Fishermen, too, cannot be always distinguished from boatmen and sea- faring workers, whilst there is even more confusion- between the sellers of milk, who are often residents of the town, and the rural dealers in cattle. Often where the head of the family is a man, the entry is of the latter, but if he be away, his wife will return herself as a milk seller, and so on. Scavenging, which figured in the returns of 1881 to a great extent under the honorific title of " municipal service" seems to have been more correctly allocated in the present tables. But agricultural labour has suffered grievously from its dangerous proximity to the class of the " general unskilled.'' On the whole, the artisan and professional classes appear to have been the most correctly returned, or the easiest to classify, though in the case of the former, the details indicate in several places the want of uniformity of return ; but their grouping is not deficient in general accuracy. The scheme of classification can be seen in operation, first, on the broader lines General of the functional Orders, in Summary Table XIII., on page xii. of the first volume of results of the Imperial Returns, amd again in detail, in Table XVI., on page 455, &c. of the same ^^ ^^°^"^ volume. In the Provincial volumes, the serial number of the latter return is XVII., tion^A^' but for the Imperial series, the delay in furnishing the returns of caste, &c. rendered it advisable to transpose the two, and to complete the first volume with -the material in hand', rather than delay its publication. In commenting on this return, it is convenient to reproduce the figures only in their relation to the general total, and to refer the reader to the original for the absolute numbers. The latter, however, will be quoted when any special group is under discussion. In Table B,at page 99, the percentage df each Order and sub-Order of occupation on the total population is shown, and to avoid the repetition of so long a statement, later on, the respective prevalence of the occupa,tion M 3 94'- Village industries. in urban and' rural tratjts is added, in the same proportional form. There^ is One feature in this last series to which attention may as -well be called at' once. This is, that though the general ratio of the urban population to the whole is about 9^ per cent., in the case of occupations, this proportion is exceeded in every one of the Orders and sub- Orders, except: those reliating to agriculture, pasture and village service. It nearly recedes to it' in" the case of general unskilled labour, which has been mentioned as probably recruited very considerably from the ranks of field and farm hands, and again amongst the potters, cane- workers, and the purveyors of; minor forest produce, such as leaves, honey, rubber, lac, and so on. This fact shows the remarkable simplicity of the industrial structure of the community in India. A little has been said on this point already, but before entering into the examination of the general return it is worth while to. dwell upon it in a little more detail. In respect to the relation of the village community to the land there is a different system to be found in nearly every. Province and large State, and in detail, too, the respective social grades are diversely arranged. But on the "Vrhole, the component parts of the society considered as an economic, unit, are wonderfully uniform through- out India. This, however, cannot be held to include Burma or several of the hill tracts, since in the latter the community is hardly a settled one, and in the other, the caste system, as has been mentioned in a former chapter, does not prevail, so the facilities for separation are greater. The marginal-table shows the relative strength of these component parts. Speaking generally. Table A. — Showing the, main' occupations the village community consists, firstly, returned. of the cultivating class, and, secondly, of those who minister to their most pres- sing wants. A cross division exists, on one side of which stand those belonging to the inner circle of the society, that is, the occupants of the land and the artisans or menials to whom tradition assigns a cognate racial origin ; and on the other, the lower village mehials, the agrestic serfs, and others ' whose occupation, in the present day, as in the time of the Chinese pilgrims, ■rele'gates them to an inferior position, and who are, no doubt, also separated by their direct relationship to the races displaced from the ownershjp of the land by successive waves of aggressive immigration from the north. The former are admitted as residents with- in the village site, but the latter must dwell without those limits, or on- a special site of their own. In the process of evolution there have been successive passages upwards from the one to the other, but the transfor- mation IS a mattpr of many genera- tions ; this is a question that will have to be treated in connection with the ethnology of the people, in a later chapter of this work. We have it then, that the backbone of the village IS the _ agriculturist, or he who holds the village lands. The labourer ' found, generally in the ra^ks the helots, above mentioned ,of others a little higher social, scale. The pastoral class be distinguished first as the With which we are not concerned, but which self chiefly with sheep and Pcr-centage of each Occupation on Occupation. Total Kuiral Poualation Population. 1. Landholders' and tenants 52-98 57-06 2. Agricultural labourers 6-50 6-92 3. General labourers , - 8-87 8-77 4. Graziers, shepherds and wool- 1-45 1-49 workers. 5. Cotton workers 3-07 2-75 6. Goldsmiths - - . 0-62 0-50 7. Blacksmiths - . - 0-55 0-52 8. Brass and coppersmiths 0-14 0-10 9. Carpenters - 1-06 0-96 10. Masons - - 0-36 0-24 11. Barbers -- - - - -. 0-90 0-89,' 12. Washermen 0^72 0-71 13. Fishermen 0-95 0-94 14. Oil pressers 0-69 ' 0-69 16. Potters 0-82 0-81 16. Village servants 1-07 1-13 17. Leather workers - , 1-14 1-09 18. Scavengers, 0-40 0-35 19. Priests ' - 0'60 0-56 20. Mendicants ' - 1-95 1-78 T.otal,! primitive 84-84 88-26 21. Milk sellers 0-35 22. Grain and pulse dealers 1-10 — 23. Shopkeepers - •'- 0-88 — 24. Money-lenders' 0-34 — 25. Grocers - , ■ * 0-74 — 26. Tailors 0-42 27. Piece-goods ,-"C; ' ■■'Rf,'.C Urban. Rural. I. ADMINISTRATION - ' ' 195 25-61 74-39 1. Civil Serviqe of the State 2. Service of Local Bodies 3. Village Service ~ - - - U. DEFENCE ^^^'^t . ^'^ . 0-83 0-04 1-08 0-23 49-51 79-15 5-03 70-88 50-49 20-85 94-97 29-12 4. Army t - - 5. Navy and Marine 0-23 70-92 48-56 29-08 51-44 III. SERVICE OE OTHER STATES 0-18 , 45 '28 54-72 6. Civil and Unspecified 7. Military - . . _ 0-13 0-5 40-78 56-67 59-22 43-33 Total, Class A, Government 2-36 31-51 68-49 IV. ]?ROVISION, &c. OF CATTLE 1-27 613 93-87 8. Cattle-breeding, &c. 9. Training and Care of Cattle, &c. 1-25 002 5-44 ' "' 46-86 94-56 53-14 V. AGRICULTURE 59-79 2-72 97-28 10. Landholders and Tenants 11. Agricultural Labour - -'^:t-^,f. 12. Growth of Special Products 13. Agricultural Supervision, Sec. 52-20 6 -.50 Oi-79 0-30 2-42 3-74 8-56- 15-81 97-58 96-26 91-44 84-19 Total, Class B. Pasture and Agriculture 61-06 2-79 97-21 VI. PERSONAL, HOUSEHOLD, AND SANITARY T SERVICE - - - -/ 14. Personal and Domestic 15. Non-domestic Entertainment 16. Sanitation , - 3-91 3-48 0-02 0-41 22-78 22-65 52 - 86 23-71 77-22 77-35 47-14 77-29 Total, Class C. Personal Services - 3-91 22-78 77-22 VII. PROVISION OF FOOD, DRINK, &c. 5-07 19-94 80-06 17. Animal Food 18. Vegetable Food 19. Drink, Condiments, &c. 1-41 1-88 1-78 16-17 24-73 17-88 ■ 83-83 75-27 82-12 VIII, LIGHT, FIRING, AND FORAGE 1-33 16-26 83-74 20. Light, &c. - - - 21. Fuel and Forage 0-76 0-47 12-90 21-62 87-10 78-38 IX. BUILDINGS 0-50 35-47 64-53 22. Building l^aterials - ... 23. Artificers in Building 0-14 0-36 24-74 39-77 75-26 60-23 X, VEHICLES AND VESSELS 05 37-72 62-28 24. Railway Plant 25. Cai^ts, &c. 26. Ships and Boats 0-01 0-02 0-02 72-90 39-57 9-56 27-10 60-43 90-44 XI. ARTICLES OF SUPPLEMENTARY .REQUIRE- 1 MENT - - 1- 27. Paper - - , 28. Books and Prints '29. Watches, &e. -» > 30. Carving and Engraving 31. Toys a.nd Curiosities 32. Music and Musical Instruments 33. Necklaces, Armlets, &c. ^ 34. Furniture . . - 35. Harness 36. Tools and Machinery 37. Arms and Ammunition 0-40 0-03 0-03 0-02 0-02 0-01 0-20 0-01 0-01 0-06 0-01 36 12 40-66 87-75 85-29 55-80 36-46 19-78 25-80 60-90 61-56 25-72 52-44 63-88 59-34 12-25 14-71 44-20 63-54 80-22 74-20 39-10 38-44 74-28 47-56 XII, TEXTILE FABRICS AND DRESS ; - 4-39 22-66 77-34 38. Wool and Fur 39. Silk 40. Cotton • -, 41. Hemp, Jute; and Coir > - - i- 42. Brass - - ... 0-21 3-07 ^■'^•'■ 0-16 0-84 15-81 ^6-46 18-92 16-66 -'••' 35-92 84-19 53-54 81-08 83-34 64-08 _. ._ ._ ._ N 2 100 Order and Sub-Order. A. Per-centage on Total Population. B. Per-centage in each Order of Urban. Rural. XIII. METALS AND PEECIOUS STONES 43. Precious Metals and Stones 44. Brass and Copper, &c. 45. Tin, Zinc, and Lead 46. Iron and Steel XIV. GLASS AND EARTHENWARE 47. Glass and ChinaTrare 48. Earthen and Stone^rare XV. WOOD, CANE, AND MATTING 49. Wood 50. Cane, Matting, and Leaves XVI. DRUGS, GUMS, &o. 51. Gums, Resins, &c. 52. Drags, Dyes, &c. XVII. 53. LEATHER, HIDES, &c. Total, Class D. Preparation and] Supply of Material Substances j , XVIII. COMMERCE 54. Money and Securities 55. General Merchandise 56. Dealing — unspecified - 57. Brokerage and Agency XIX. TRANSPORT AND STORAGE 58. Railway 59. Road - 60. Water 61. Messages 62. Storage and Weighing Total, Class E. Commerce Transport, &c. and XX. LEARNED AND ARTISTIC PROFESSIONS 63. Religion 64. Education 65. Literature 66. Law 67. Medicine 68. Engineering and Survey 69. Other Sciences 70. Pictorial Art and Sculpture 71. Music, Acting, and Dancing XXI. SPORT AND GAMES, &c. 72. Sport 73. Games and Exhibitions Total, Class P. Professional XXII. 74. EARTHWORK AND GENERAL LABOUR , XXIII. 75. INDEFINITE AND DISREPUTABLE "I MEANS OF SUBSISTENCE | XXIV. INDEPENDENT OF LABOUR 76. Property or Alms 77. Supported by the State Total, Class G. Indefinite and In- dependent 1-33 0-62 0-14 0-02 0-55 0-82 0-82 1-50 1-00 0'50 014 0-03 0-n 114 15-43 1-63 0-39 0-41 0-67 0-06 1-38 0-10 0-63 0'28 0-11 0-26 2-91 197 1-11 0-17 0-10 0-08 0-18 003 0-07 0-01 0-22 005 0-02 0-03 2-02 8-87 0-54 1-66 1-50 O'lG 1-66 23-46 26-96 35-41 B6M2 14-80 10-60 70-72 10-23 15 18 17-21 11-08 ■24-91 9-89 28-53 13 97 20-32 30 42 27-91 27-56 28-36 52-39 30-16 55-93 25-03 28-28 38-19 31-21 30-30 26-56 20-05 33-11 53-98 62-56 29-69 50-55 20-76 52-31 23-57 18-19 19-79 17-48 26-35 10-46 16-48 19-69 15-90 64-72 1210 [-76-64 73-04 64-59 33-88 85-20 89-40 29-28 89-77 84-82 82- 7S 88-92 76-09 90-11 71-47 86 03 79-68 69-58 72-09 72-44 71-64 47-61 69-84 44-07 74-97 71-72 61-81 6S-7U 69-70 73-44 79-95 66-89 46-02 37-44 70-31 49-45 79-24 47-69 76-43 81-81 80-21 82-52 73-65 89-54 84-52 80-31 84-10 45-28 87-90 101 Order III. — Service of other States. — This head was intended for the persons employed in one State but who happened to be in another at the census. It has been construed however, in some cases, as including those of the State furnishing the return. 6. Civil officers 7. Military, &(!. Total, Foreign States 358,079 141,951 500,030 8. (a) Horses and homed cattle 2,950,649 8. (b) Other draught animals 33,067 8. (c) Sheep, goats, pigs, &c. - 971,682 Total, 8 - 3,585,398 9. (a) Training and care of cattle ... 9. (6) Vermin catchers, &c. Total, 9 Total, Pastoral 54,467 5,984 60,451 3,645,849 10. Landholders and tenants 11. Agricultural labour 12. Growth of special pro ducts and of trees 13. Agricultural supervision &c. 149,931,159 18,673,206 2,261,481 Order IV. — Fastoral Occupations, Sfc. — It was found advisable to considerably [ subdivide this order. First, the number of those concerned with horses and horned cattle had to be distinguished from those who dealt with smaller cattle, who form a totally different class of the community. Then an intermediate group was suggested for the breeders and catchers of other draught animals, the use of which is practically confined to certain parts of the country. Camels, for instance, are principally to be found in the dry plains of the west. Elephants on the other hand, haunt the moist jungle of Assam, Mysore and Burma. Asses, as has been already mentioned, are the adjunct of the humble potter and washerman, and mules appertain chiefly to the military establishments of artillery and commissariat. To the number of those occupied in the breeding of small cattle are to be added most of the blanket weavers, now under the head of textiles, as the occupations are exercised by the same class. Lastly, there are the farriers, veterinary surgeons, and so on, with a small contingent, under a separate head, of those who live by catching monkeys, crocodiles, and other noxious animals. Order V. — Agriculture. — Of this great class the subdivision is all that needs comment here. The return includes, under the first head, 97,674,965 tenants holding under an intermediary, and 51,592,844 occupants holding under the State, but there is some reason to suppose that in Peuda-' tory territory these terms are not uniformly applied. Under the second head we find 4,699,897 supported by farm labour, that is, by field work of a more or less per- manent character, as by the year or for life. There are, moreover, 13,973,309 given as ordinary field labourers, a number which, as before mentioned, has to be supplemented by perhaps as many as 18,000,000 more, out of the class entitled general unskilled labour, that comes towards the end of the return. The third subdivision comprises a considerable number of persons who ought to appear under the first, such as market-gardeners and those who live by growing spices, betel, areka, and so on. The 560,000 planters, too, of tea, coffee, and indigo, with their staff and labourers are here included. As to the care of trees, which forms a part of this section; the main items are the employes of the Forest Depart- ment of the State. The last subdivision consists chieflv of the agents and managers of estates, who are most numerous in Bengal and Madras, where the permanent settlement of a great part of the province has stimulated the growth of absentee landlords. Order VI. — Personal, Household, and Sanitary Services. — The main items of this group have been already mentioned in various parts of the introductory remarks to this chapter. We have, in addition to the Barber and Washerman, 2,492,544 indoor servants, and 1,241,521 persons living by service, but of what description is not specified. Water- carriers support some 900,000 persons, and cooks a little over a quarter of a million. Some of the latter are no doubt keepers of cookshops in the larger towns, especially in I 869,544 Total, Agriculture 171,735,390 14. Personal and services 15. Ifon-domestic ment Sanitation domestic 1 entertain- \ 16. Total, Personal, &c., Ser vices :} 10,008,387 47,803 1,163,882 11,220,073 AT Q m102 tlie nortt-west of India, wteye thgre is a large «\papulation so addicted to officinal purity >tliat ttey cannot takje advantage of such services. The second group under this order, too, is a very restricted one, and is largely recruited from the north. The mass of the people require no inn pr refreshm^nt-i^oom when they go on a journey. The^ bathe in the nearest pond Or riy;er, and sleep on a chance verandah, or in the village oi" town meeting-shed. In the north of the country, where life and property is supposed to be in greater danger of attack, and distances are longer, the rest-house or Sarai is a recognised institution, , and the bath-house introduced by the Mu^alman, is found in every large toiful though not 'perhaps in the perfection of luxilry described by Lady Mary Wortley Montague. 'Sanitation, filially, is an item divided between the scaven- gers', who Are, as above .stated, a village institution, and those who i. supervise itl^eir labours in the towns. ' , . Order YII. — Food, 5,133^315 Total - 14,575,593 of regetable food is, as is to be expected, ihe llargest of the three. The chief items have been discussed already. Hext to the grain and pulse dealers come the grain ..parchers, an important section^' of the community in towns, who return 795,177 persons. Sweetmeat selling is an allied occupation, and comes after the former in numerical strength. The sellers of fruit and vegetables, who are mixed up between this class and that of agriculture, are the only others that need be mentioned. In' the matter of drinks and condiments we have a very varied selection. The grocer holds the first place with over two millions depending- on his business. The consumption of betel leaf, areka nut, and toba(Cco is .almost, universal,: but > the persons who subsist by the exclusive sale of these products number only just over the million. Many of the first named, however, are grow:ers as well as sellers, and are therefore to be found amongst the landholders. Throughout the coast, distrjiots, and in many inland tracts, the juice of one or other of the palm tribei of trees is habitually drunk by all the middle and lower classes, and the class i known by the general term of Toddy-drawers is returned at nearly a million. The distiller and vendor of country-made spirits, who comes from the same section of the community, where he is not a Parsi, is returned at only a third of the above number. The number of those depending on the sale of opium for their living is remarkably few, considering the wide distribution of the classes that use the drug. It may be borne in, mind, however, that in British territory the 'sale of opium is restricted to shops licensed in the larger towns, whilst in Rajputana, where every house of good social position amongst the dpminant tribes begins, the day with an aubaine of Kiisumba, th^ decoction is prepared like any other po];'tion of the family food, so that no special occupation is attached, to its manufacture or sale. The small return of the purveyors of preparaitions of hemp and their depQndenjbs, amounting to 17,881 only, cannot be so explained, but no doubt a good many are to be found in the category of Farmers, in Order XYIII. In fact, the whole of this group suffers from the confusion between the dealing in a special product and dealing in general. Order YIII. — Light, Fuel and Forage. — The first of the two subdivisions ' is mainly occupied by the oil presser and his companions, under various titles. ■ Gas and candles, pnly provide for some 1,233 of the lieges, but there is an abundance, vald4 deflenda, of ,the dealers in the mineral oil that is expel- ling the native product from the market. We have, however, a trace of primitive custom in the 4,060 people who trust for their bread to the manufacture of torches useful though malodorous, on a dark journey; ; In the other section of this Order, forage and fuel are combined in the item which furnishes the bulk of the persons' concerned because the class that briiigs in grass for sale is equally disposed, according to the season, to provide firewood, the main object being to get a head-load to the market. The prevalence of the use ,of jupwdnng fuel has beenmentionedrabove." As this is a "pure 20. Light - - 2,163,593 21. Fuel and forage - - 1,358,664 Total - - 3,522 357 103 product, eeremonially speakin^er, roost families prepare their own cakes for burning or selling, so the occupation is not specialised. But near a town tlie cowkeepers seem to have a monopoly, or nearly so, and thei.south walls of the houses in their Ward. are often decorated all over with r9ws of these solid brown wafers, smacked on to dry. On the approach to the TdJ at Agra, there is usually a peculiarly extensive display of this. sort. : OUd'er IX. — Buildings. — The first section under this head is a very small one, ' ' since roost of the brick burners, are returned as potters, from their main busijiess, and also from'the habit of returning lhe caste for the occupation. The lime burnero, too, do not coiifine' their labours to this commodity, but combine it With saltpetre digging and other cognate trades. In the second section the and stone cutters fill neai'ly the whole tale. The number at 17,678, as the majority are, in the the villages, merely ^. Building materials - 23i Artificers in building - 4i 0,142 1,027,597 Total, Building ' - 1,437,739 masons, mud- wall makers, of builders is understated carpenters. 24! Railway plant 25. Carts, &c. - 26. Boats and ships Order X. — Vehicles and Vessels. — ^We have here a still smaller class to deal with, as the railwa,y establishments are confined to a few large towns, and some of the original work is brought out from Europe, and a good deal of the rest is done by people who return their general occupation as blacksmith or carpenter, not the special branch of it. The cart-inakers are merely -a branch of the carpen- ters, and are not specially devoted to the manu- facture of articles for which the demand in their own village is small, though, in the aggrgga,te, considerable. Much the same may be said of the boat-builders, but the canoe-digffers, where those vessels are in fashion, are duly returned as such. Total, Vehicles, &c. 42.529^ 47,728. 56,261 146,508 Order XI. — Articles of supplementa/ry requiremmt. — This rather clumsy title includes what is, for the most part, in demand after the first, wants of social man . have been satisfied. . .Paper, to take the first item, is made to a considerable extent in India, on two systems. In a small way, the native hand- worker holds his own, but his occupation is restricted to comparatively few places and families. The larger operator uses machi- nery, and his' emplbyes are probably returned under - the . head of .labourers, to some extent. The largest works are near .Calcutta, and from them, it may be remarked, en passant, most of the paper, amjounting to about 290 tons, more or less, was obtained for the census schedules. In Grwalior, too, where there is abundance of the speciarquality of gtass required for tough paper, .mills have been erected. There are others in Poona and Lubknow. To pass onto books, we may note that with a 27. Paper * . - 78,153 28. Books, &c. - - - 94,277 29. Watches, &c. 11,638 30. Carving and engraving 44,336 31. Toys, &c. - 69,747 32. Music and musical instru- 1 ments - - - j 21 397 33. Armlets, necklaces, &c. 586,'!;39 34. Furniture - - - 15,616 .35. Harness - - 17,406 36. Tools and machinery 173,593; 37. Arms and ammunitiqn 42,365 Total, Supplementary 1,155,267 population of whom oiily six in a hundred can read and write, authorship is not likely ^ ^- " ■ -■ -- — Qf livelih-ood; so the amount of work thrown on 12,000 people. Watchmaking, again, is not yet the press is exe'cuted by about 12,000 people. W atcn-maiang, again, an acclimatised industry, as the mass of the population require no more accurate timepiece than the sun, witli whose movements' they are remarkably familiar. In the carVinff and engraving line it seems that there are a good many carpenters and turners who have preferred the general title. In the next section, the toy makers, especially those who deal in kites and clay dolls, or images of the more popular deities, are found' in towns generally, but in the country, only in certain parts of India. Possibly kites- are often hbme-made, otherwise it is not easy to account for the numerical deficiency of those who further so very popular a recreation, and one, too, that has certainly m some tracts if not universally, a religious sanction. Next to the kite-naaker, the maker of pipe-s'teras is most numerous. This, too, i s somewhat of a local industry, as the north patronises it in preference to the shorter smoke of the centre and south. The bowl, it must be remembered, is not made by' the same person. In regard to musie, it is N 4 104 as well to point out that, though, in the upper classes a stringed form of instrument may be the favourite, the national instrument of India is the drum, and its usual accompaniment is a long brass horn. As a rule, the performers make their own drums, and will appear, therefore, either in Sub- order 71, or, as they belong generally to the village menial class, some of them are sheltered under that title. The next group is the largest in the Order. The makers of armlets are not accurately distinguished according to the material of which their wares are composed, as was intended, but they form in the aggregate the main portion of this section. Beadmakers are closely allied in caste to the workers in glass bangles. Rosaries, or to speak more correctly, chaplets, and flower garlands, are the articles next in favour under this head, and there are 8,007 people returned as dependent on the fitting on of armlets, a process of some nicety. Furniture is not a section calling for remark. The chief item in the house of the masses is the bedstead, which does not appear in this group, as its legs are turned by a carpenter, and it is strung by another artificer. Outside offices, tables are required only by the literate minority, and chairs, though more common, perhaps, are not a regular article of use in any household native to the country. The report as to the use of these articles by the inhabitants of Kafiristdn requires confirmation. Harness making is probably carried on to a considerable extent by the leather workers, and only the embroidery of the saddle-cloth and the, adornment of the cotton or woollen reins fall into this section. The next section, referring to tools and machinery, is shared with the carpenters, who make the ploughs and also part of the looms, and these are the main items here. The knife grinder is an itinerant, as in England, and when he settles, he adopts a fresh trade, such as that of blacksmith, or, in a small way, cutler. The makers of sugar presses, too, come into this group. The carpenter makes these also, so it is probable that the 1,149 under this head are concerned in the production of the iron press, which has come into fashion during the last 10 or 12 years. It has been said that in some parts of India the adoption of this form of press, which gets more but impurer juice out of the cane, is the prelude to the use of mineral oil in the house. The first step on the path of innovation leads to others. The last subdivision of the Order is that of the provision of arms and ammunition. There are only two items under this head that require mention. First, the manufacture of powder for blasting purposes, which is carried on very sparsely, and, next, the preparation of fireworks, which, as conducted in India, is a higher branch of the same trade. Fireworks are an, accompaniment of most festivals, domestic or public, but the articles are usually of a simple character, though there are masters of the craft in some of the capitals of native States who enjoy a great and deserved reputation for their skill. The swordmaker and those who provide the forest tribes and the : watchmen with bows and arrows are in a majority in the native States, or possibly the latter trade is better regarded there than under British rule. Ordek XII. — Textile Fabrics cmd Dress. — This large group is subdivided according to the material used. The woollen workers consist mainly of the blanket weavers and wool spinners, who number 411,269, and belong to the pastoral class. There is a largish item, too, of cloth sellers, which means, . in all probability, sellers of cotton piece goods, ' as the vernacular term, may apply to both. Shawl weavers come next, with a dependent population of 41,831. Of these 30,000 are in Kashmir and over 9,000 in the Panjab, where Amritsar has secured a regular colony of both shawl and carpet maker?. The section dealing Avith the silk workers contains two very diverse elements. First, there are some 87,400 silkworm rearers and cocoon collectors. The latter are members of the forest tribes who make it their business to. look for the Tassar moth and bring in the cocoons to the dealers in the plains. Most of the former are in Bengal and the rest in Burma, Mysore, and Assam. In the first named they are located chiefly in the Murshidabad, Rdjshahi, and Mdldah districts. In the other provinces they are probably collectors. Then come the workers in silk. These constitute an honourable offshoot of the great weaver class. The differentiation must have taken place many generations back, as they are mentioned with commendation for their piety towards the sun and the general respectability of their behaviour, in an ancient inscription of the time of Kumdra Gupta in Mandesur, a town of Central India. In the extreme south-east of the Madras Presidency they, form an independent community, who have within the last 38. 39. 40. 41. 42. "Wool and fur Silk • - Cotton Jute, hemp, &c. Dress, &c. Total, Textiles, &c. - 587,701 319,397 8,820,466 461,193 2,422,510 12,611,267 ±05 tew generations establislied a priesthood of their own, in defiance of the Brahman of the neighbourhood: The census return shows only 204,000 of them, however, and it may be that some have been returned as embroiderers. Apart from articles made entirely of silk, they are much engaged in working at the edging of -cotton garments of the better sort and in braid. The cotton industry has been already cliaraoterised as a village occupation, but there are 1,668,895 persons in towns who return it. The main body are found under the head of spinners or weavers, the distinction not being well drawn. These, with the factory hands,- reach a total of 7,380,278, There are also the calenderers, printers, and dyers, numbering nearly half a million. One of the qUiegtions of the day is whether importation and the establishment of machine-driven cotton-works in the larger towns, especially in Bombay, has not tended to drive the industry out of the hands of the home-worker into those of the general labourer, who soon picks up one of the special branches of mill- work, on the one side, and into those of the foreign operative on the other. It is impossible for the census to adequately deal with this. As to the first point, however, ifc seems both from the census and the trade returns that not more than 120,000 hands, maintaining, say, 600,000 persons altogether, were employed in the factories in 1891. The competition of the European operative is, of course, more serious than that of an infant enterprise. Twist, as we have seen, is more than holding its own, but this does not afiect the village weaver. The question is, what room for his wares is left by the Ex. 23,750,000 worth of piece goods annually imported, on the average, during the last decade ? The only provincial Eeport on the census in which this question is dealt with is that ol. Mr, Stuart, for Madras. He shows that the occupation has slightly increased in the, last 10 years, and by comparison with the caste-return, he finds reason to believe that it is recruited to some extent from those beyond the pale of the castes with whom the function is traditional. He assumes, accordingly, that there is something to be gained by engaging in this pursuit, though the improvement is but slight. On the other hand, if wo recognise the chances of the entry of some weavers under a more general title, as is" not uncommon with Musalmans in the north, and the difl&culty ej^peuienced by an artisan, especially one of so technical a class as the weaver and a member of a community, to boot, with whom occupation is generally hereditary, in. changing his trade, it is difficult to accept the numerical test as entirely disposing of the question. There is this to be said, that the competition between him and the foreigner is less than the above figures might seem to denote. For the home product, strong and coarse, but genuine and durable, still holds the agricultural market amongst the lower classes, and, indeed, amongst the raiats generally, a fact that is obvious to any one who goes about amongst the country people. In the towns, the competition is beyond a doubt severe, for the finer fabrics of the home-loom must be more expensive than the machine-made product of Europe, now that processes are so much improved and freights so low, and it is the townsman that takes first to a novelty. On the whole, therefore, looking to the very sniall proportion of the urban population, and of the wearers of fine raiment amongst the rustics, it does not appear that the field of the ordinary weaver is yet usurped by any competitor. Tlie manufacturer of muslin and " woven-air " and so on, is the one to sufier. The next sub-order is a composite one. The persons depending on jute production and working are mostly to be found in Bengal, where the industry supports, or helps to support, a good many more than the 120,000 or so who have returned it as their special industry. It has been said in the last chapter that on the whole this occupation is a very thriving one, though perhaps in the present season it may have been temporarily depressed. Along the west coast as far north as Bombay, the preparation and manufacture of coir, or the fibre of the husk of the cocoa-nut is a recognised industry amongst the fishing and other • classes of similar position, but it is returned as the sole or principal means of livelihood by only 108,169 persons. The ropemakers and workers in hemp and flax are nowhere devoted solely to this occupation, but undertake it as an adjunct to village service or some other means of subsistence. This accounts for the fact of the return being limited to 192,000 people. Finally we have the tailors and their comrades to consider. The former seem to have originated from the calenderer or weaver class. Many are mere dealers in cotton piece-goods rather than actually makers of clothes. Thus there are but 1,209,343 dependents on tailoring, and 915,796 vendors of piece-goods in the return. But to the former there may be added some 160,000 who return special branches of the trade, such as cap, bodice, or sari-making. Turban-binders again are an offshoot from the sartorial mystery. Embroiderers, on the other hand, form a distinct class allied to the finer weaving industries. There are returned of them only 44,548, so elsewhere but in the great r 78388. 43. 44. 45. 46. Gold Brass Tin, Iron , silver, and stone.'i and copper &c. and steel 1,783,874 405,600. 59,048 1,572,911 Total, Metals, &c. 3821,433 106 towns of Lhe north, they liavo evidently given some other title. Lastly, we may throw in a word for the makers of umbrellas, since they, like the weavers, have probably suffered from forejign competition. The indigenous, article of wa^ed calico is being displaced by the. more durable gingham of Europe., and it is only alopg the coast^, and where there is the heaviest rainfall that the clumsy .but efficient arrangement q| leaves on bamboo framework holds its own. Like brass or copper pots the imported umbrella is usually taken to be a symptom of improved circumstances. Order XIII. — -Metals and Precious Stones.— 'Nearly the whole of the first sub- division is absorbed by those who live by working on personal ornament's, for the ' numlaer assigned to" assay ers should, liodOubt, com6 under th(^ general heading of Groldsmith. Of the rest we need only draw attentibh to the inanufatcture of gold wire, lacb, and t)raid, numbering 47,000, and the 23,676 persons living by washing out tlie refuse of the goldsmith's shop, and recovering from it the minute particles of the precious metal that have been clipped or filed off in the course of the day '6 work. Of the precious stone dealers, only those in pearls and rubies, in both of whom divers and miners are probably included, show any significant figure in the return. Coral dealers, like those in pearls, are most numerous in the Madras Presidency, where the fisheries are situated. Those persons who come under this head on the west coast are probably connected with the Bahrein fisheries in the Persian Gulf, residing in the Indus delta, at Tatta or Karachi. The brass and copper workers must be taken in the lump, as the detailed heads are probably not mutually exclusive, and a large proportion of the entries do not specify any particular branch of these industries. Tin-working is an important town industry, though but sparsely represented here. Some of the workers in this metal are to be traced into the heading of Lampmakers, for, since the intro- duction of mineral oil, the tall and graceful brass lamp has given place to the cheap tin product of the locail bazar. The cans, tod, in which this oil is brought into India are much appl-ebiated by the people, owing to the varied uses to which they can be put after a little maiiipulation by the tinman. For instance, probably balf the watering-pots in the country consist of one of these cans, stiffened with a rough wooden bar or two, and adorned with a i-ope or wire handle. The supplies of vegetable oil and even of clarified butter, or ghi, that have to be sent from the market town to the city, are put into the dead shells of their rival. In architecture, too, the can has introduced a new feature, surpassing even the corrugated iron plate in unsightliness^ for it is flattened out for roofing, tnade into spouts, or arranged along the pedittient. As regards iron -workers, enough has been said in connection with thd village organisation. With the exception of the smelters, who form a small section of the hill population in the Panjab and other parts of India, the same person combines all dealings with this metal, so the sub-divisions shown in the return are but partially representative. Order XIV. — Glass, Earthenware, Sfc. — The workers in glass are prone to return themselves under the title of bead or armlet makers, so very few come under the head- ing appropriated to the material in which they work. The second sub-division of this order is composed of two main pursuits ; first, the potter, who has been described above, and then the makers and rougheners of the hand-mills used in every household in the preparation of meal, and of the flat stones for the mixing of the condimeiits that go to flavour it. Order 'KY.—Wood, Gone, and Matting. — In the first section here there are three main occupations. First, the carpenter, then the sawyer, his inferior, and lastly, the man who deals in timber and bamboos. As to the last item, it must be understood that he is altogether different from the worker in cane,, who comes in the second section. In towns, 47. Glass and china - 14,419 48. Earthen and stoneware- 2,346,204 Total Glass, &c. - 2,360,623 49. Wood - - 2,868,433 50. Cane, &c. - - 1,424,579 Total - - 4,293,012 the occupation of providing wood and the bamboos that are so much used as lathes for rooting is quite a special one, and it was 107 -' 'Or1)er XVL- -^Drugs, Gums, Dj 51.' Ijuths, &c. .52. .Drugs, dyes, &e. ■ ■"' Total (•■■ ■ U' : 76,18.6 . 315,389 - ?«,575 desirable, therefore, to try and distinguish it. ^^he cane- worker, under a variety of *™«^ IS generally, as he is, or was, in England, a vagrant of shady repute. Some inquirers have sought to identify him with the gipsy of Europe, and quote Changer, the i'anjab variety, as the origin of Zingari, Tsing^ne, and so on, as well as D6m, who performs similar functions further' to the east, and is, they say, fhe Il6m of Borrow, Leland, and other gipsyologists. This, however, is a matter to be considered later. In this group there IS one occupation which we must dissociate from its companions. This IS the manufacture of leaf -plates, which is the function of a special class in Bengal and Bomba.y., The purveyors of these articles, which, are in much request at qgiSte-leasts and thp like ceremonies, are not nomads, and in the west of India they cguabiiiie with the above occupationj ihat of temple-ministrants. Byes, Sfc. — The first section is confined to the collectors or preparers of minor forest produce, of which the chief representatives are lac and catechu, or cutch, as it is called in Burma, where, and in Bengal, the industry seems most to flourish. In Burma, however, it is combined with wood-cutting, so the number ^. ' ' is thereby reduced in the return. The central hills of Bengal and the North- West and Central Provinces yield most of the lac- workers. There are also re^iirned 26,144 collectors of honey and similar wild produce. In the 'category of drug and dye sellers, the perfumer comes next to the saltpetre digger "^and the chemist, or rather, the person who deals in the ordinary village medicines. This last represents abqut 14,000 families, which seem hardly sufficient to supply the whole demand for opium throughout the country, as has been suggested, eV^n if we grant that aii;;^ of them, except the few in the largest towns,' know anything of pharmacy in the'^^English sense of the , term. Some of the dealers in dyes ought clearly to have been' .classed in'brder XII., as they probably use the dye, and do not merely prepare or sell it. Amongst the miuor items in this class it is not altogether satisfactory to find about 300 dealers in aniline dyes returning themselves as such, without shame or fear. In Persia, it is said, the trade amounts to felony. , Oeder XVII. — Leather, ^a. — The first four items in this group, which amount to 3,131,156 persons, represent the members of the village staff, as already described, who have extended their original functions so as to include more than the mere tanning of the hides of the dead cattle of the ueighbourhood. The enlistment of a good body of this, class into the factories established in Cawnpore on the European system has been already mentioned. The demand for hides, horns, and other exuvice of the saorisd animal has increased considerably of late. In 1881-82 the number of hides and skins exported was 24,802,043, and in 1890-91 it was 32,742,431. The value o-f the horns for the two years was respectively Rs. 1,329,150 and Es. 2,077,190. Bones, again, were exported to , the value of Rs. 1,327,500 in 1887-88, and of Rs. 2,309,570 last year. All this is favourable to the excheiquer of the village tanner, but as to the last, the benefit to the country at large is less than doubtful, since the soil of India thereby loses hearly 44,000 tons of fertilising material which it cannot spare. Oeper XVIII. — Gommerce. — Dealings in money are naturally the subject with which this order opens. Taking the bankers 53. .Leather, &c. 3,285,307 5.4. Money, &c. - 55. General meijchandise . 46. -.Dealing, unspecified 57 (a.) Brokers and agents 57 (6.) Contractors and farmei's Wotal, Middlemen, SfC. - . Total, Commerce 1,128,283 1,186,892 1,907,555 259>726 203,123 462,849 ,4,6,85,579 and money-lenders together, the persons sup- ported by them is very nearly a million. It should be understood that in the term money- lender, we include that of pawnbroker, for the business done in the village is largely in advances on ornaments pledged.^ , The only other considerable item in this section is that of; money-changer and tester, with 93,233 against it. The function is . often mixed up ' with lending, so the above figure is probably below the reality. General merchandise, ' afir^^e have seen;,, includes many of the village shops which deal in a variety of articles i^.gsn^arl demand, - Therf are 1,494,165 persons set forth against the term, unqualified b^ further' description, and 1,028^980 come under the head of general merchants. -Hawkers and -pedlars ai!e mndereitated at 14^000, as many purvey food; others are metrejy the 2 108 The second group relates, or was intended the business and live on profits, not merely not accurately distributed between the are 58 (a.) Eailway administration 58 (6.) ,, working staff Total, Railway 59 (a.) Carts, &c. 59 (6.) Pack aniirials Total, Road 60 (a.) Ship and boat ownei-s 60 (6.) Tailors, &c. Total, Water 61 (a.) Post, &c. 61 (6.) Telegraph, &c. Total, Messsngers, Sfs. 62. Storage, &c. Total, Order 29,169 231,040 285,187 1,126,847 676,339 i,8o3,i86 ,92,701 697,051 789,752 307,903 19,171 027,074 747,794 3)952,993 travelling members of a shop in the market town, and return themselves there. The sub-order relating to middlemen, such as brokers and the like, has been split up into two. The first, which includes mere general intermediaries,, shows 240,592. broke;rs and agents, and 19,000 auctioneers, &c. to relate, to those who take a share in on commission. The numbers in it items, and there are too few farmers with an excess of contractors not otherwise specified. Order XIK.—Trans'port and Storagc-^B-Qve it appears that the railway, in its capacity of carrier, and irrespective of mechanics, feeds some 285,000 people, mostly, of cou,rse, in, the, lower ranks of employ, such as porters, pointsmen, sig- nallers, clerks, and so on. The next section is a large one. The persons connected with carts are almost entirely of the agri- cultural class, whilst the bearers of palki, or palanquins, a mode of conveyance rapidly falling into disuse, hail from amongst- the fishers. It is a common practice for both to engage in their ordinary occupation for most of the, year, and take their, cart and bullocks, or, their thews and sinews, as the case may be, to some large town during the agricultural recess. The pack-animal drivers are very widely distributed, but, as a rule, the bullock traffic that used to be so celebrated is now confined very much to Rajputana and Central India. The camel driver, who is also the breeder of the animal, flourishes all over the west of India, north of the Gujarath division of Bombay, and the elephant is mostly found in this capacity in Assam and Burma. The third section, which contains those who go down to the sea in ships, is shared by the return of fishers, as above stated. In other respects no comment seems necessary. It is much the same with the sub-order relating to the conveyance of inessages, which is sub-divided into telegraphy, &c., and postal and others. The unclassified is liere, again, the largest item. In the last section of the order, we have the class of worker attached mainly to grain and other produce warehouses, and thus differing to some extent from the ordinary unskilled labourer. The weighmen, also, are in the same position, and these two cover nearly the whole field of the sub- order. now reached an order where the range, of interpretation is at its maximum, and every single sub-division seems to require a goo4 deal of explanation. The term Profession is held to include both the liberal and the artistic branches, and each of them, in India at least, contains a very hetero- geneous collection of professors. Then, too, it must be borne in mind that most professions are of comparatively recent development in India, always excepting that of religion. Education was confined, till the advent of British rule, to a small class of Brahmans, and during the Musal- man period, to a still smaller group of hereditary clerks. Literature, therefore, was equally circumscribed, and restricted chiefly to . theology, mythology, astrology, and the chronicle of the deeds of chiefs and their ancestors, which went by the name of history. Law, as a profession, was practically unknown, though there Order XX. — Professions.— -We have 63 (a.) Priests, &c. 63 (6.) Subsidiary religious services •■ 1,721,487 1,463,273 Total, Religion 3,184,760 64. Education 65. Literature 66. Law 67 (a.) Medical practice, &c. 486,497 280,705 226,163 469,895 67 (6.) Subsidiary medical services 44,179 Total, Medicine 514,074 68 (a.) Engineers, Ac. 68 (6.) Draughtsmen, surveyors, &c. 22,217 72,053 Total, Engineering, Sfc. 94,270 69 (a.) Astronomers, botanists, &c. - 1,354 69 (6.) Astrologers, &c. 198,656 Total, Other Sciences 200,010 70. Painters, &c. 35,788 71. Music and dancing, &c. 649,926 Total, Profes6ioa.ali &c. 5,672,191 were learned expounders .of the hieratic r systems of Brahmanic origin and of the Kuran and its cpmmentatiors, as we find is the case in the Bajputa;iia of the, present 4?ijr Medicine was in /the hands of the barber-surgeon, and his wife, with a sprinkling of herbalists and a larger throng of the possessors of charms, . speUp, , and amulets. Jn the towns could be found, no doubt, some of those who found their system on the Vedas, or, rather, the later Brahmanic literature of the orthodox period, and others Qwning to the name of Ionian (Yundni) physicians, to whom some of the present-day tdudaiores temporis acfi ascribe a wonderful and elaborate scientific body of doctrine, apparently not far below that of which we read in certain plays of Moli^re. Architects and engineers of indigenous birth and training were probably as few as they would be now, w'ere it not for the institutions established by the British Grovemnient for their education. Of pictorial art there is little to be said, and of music and the drama the less said, perhaps, the better. , We may begin the comments on this class with the occupations connected with Religion religion, which, as shown in the marginal table, comprises considerably more than hajf the prq^fessional class. The two main items, priests and religious mendicants, are hardly to be distinguished, except in the case of foreigners and Buddhists. In the Brahmanic system, the person actually officiating in a temple is a group compara- tively sparse, and lower in rank than the spiritual adviser and reciter of texts, whilst charity to a Brahman is a dntf imposed on all alike. Amongst the Musalmans, too, the extension of religious asceticism is one of the peculiarities of the faith that has sprung up in India. In Burma the distinction was drawn between Phongi, persons actually received into a religious order, and UpAzin, or novices, with a further division of the unaffiliated mendicants, who are only bound by vows of poverty. There is a considerable number shown under the subsidiary services of religion, as is to be expected from the addiction of the masses to pilgrimage and the adoration of relics or the images at particular shrines. Of these functionaries there are several different grades, as there are of the reciters of the semi -theological poems, but it will be out of place to enter into detail here, where only the general class of occupation is in question. The section connected with education comes next in serial order, though owing Education to the prevalence of the herbalist, it is numerically inferior to the sub-order of medicine, and the drummer and dancer exalt over it the category of music, &c. It contains but one large item, that of teachers and their families, amounting to 476,216 persons. This includes, of course, the village school as well as the college, since any distinction between the two can be better obtained from the speeia.1 returns of public instruction than from a census schedule. Literature is not a large class, and even as it is, the number is swelled unduly by Literature, the addition of all who returned their occupation under the general term of writing, by which they mean the traditional pursuit of their caste in some cases, and some post of a clerical nature in the rest. With so restricted a public as is before them in India, the chance of making a living by literature occurs but to few. The law, too, though largely recruited from the colleges in the present genera- Law. tion is not as well represented as we should expect from the continual complaint of increasing litigation and, of the British exaltation of the toga over arms, through which — ■ ^ , . " Encounters at the bar, Are braver now than those in war." There are nearly as many unaccredited practitioners as there are of the orthodox. The petition writer, for instance, on whose labour 29,348 depend, is often the standing counsel of the poorer litigant for the whole street. The Kazi, who is here entered to the number of 13,389, is in reality far more numerous, but as his functions are scarcely separated from those of religion in the law he professes, part of the class is probably to be looked for in Sub-Order 63 (a). Of medicine and its votaries we have already spoken. The bulk of the section is Medicine, comprised under the head of Practitioners without diploma, a conveniently compre- hensive title, represented by no less than 353,295 persons. The midwives and their families come next with 82,589. The superior class of practitioner supports only 16 494, but, no doubt, looking to the tone of recent representations on the subject, tij^j^ iiust'jD.e several fairnlies of ..this jort which are returned under the title of ■ ■ '-■■■■'■ " -•' . ^ • - 6 3 '^ ■• ■ ■ ' no Engineering, architecture, and sur- veying. Other sciences. Painters, &c Music and the drama. military- officers.-- • The last ^-section' of this sub-Order includes the" comparatively numerous and worthy idlass of hospital assistants and Other attendants at thfe dispensaries and other institution's of the class which haye sprung into existence vdth British'rulfe^ in both Crown aiid Feudatory territory. ' , Engineering and purvey are titles that explain themselves. The section comprisei* mainly, no doiibt, the persons in the employ of the State or Local bodies in ttig capacity, with their establishments. For instance, the world-famed Trigonometrical Survey has'urider it a considerable iiumber of employes wKo are far above the rank ot mere labourers, though conning below that of surveyors. There are, again, the Revenue Survey "of each' of the larger provinces, the Archseological l)epartmeiit, and several others. ,'' :"'/'', "' ,',.,,,.. .';,/. The last section of the learned professional class is that of the branches which ar© represented very sparsely in India, save in one direction. The astronomer, indeed, is to be counted toy units,'but of the domestic diviner there are 198,656. ' But a good portion; of'thelattei' is uo" doubt attributable;' ndtxo'prophecy', but "to the "pedigree- keeper, a man of considerable merit and dignity, especially in Native States of the west of India, where family is all important. The 30,056 painters returned in the next section are far from being of Academy*; rank, and many are probably artists in house- wails rather than in more moveable, material. The photographer, as is only to be expected, has invaded India with some profit, and his art supports nearly 3,000 persons. We have, finally ^ the large body of musicians and dancers. Of the former thefe are returned 372,561, and of these most are, no doubt, village performers on the drum and certain wind instruments. The dramatic arts are mixed up with other professions, and a good many of those who exercise them are entered under some euphemistic title of the less reputable branch of their functions. It is impossible to separate the two in the returns, any more than is done in the course of the education of the class in question. The Chinese pilgrims noticed the combination, and in their day^ the singei' and courtezan were compelled to reside outside the village walls, along with the fisher- man and scavenger. History seems to indicate that she was not kept out long, at least when the village grew into a town. As far as the dramatic part- of the occupation goes, however, there are numerous degrees of proficiency distinguished in practice, though not in the returns, in which last the; general notion of the employment is that expressed in one of the schedules from a town in the north, as " singing and enjoying sensual pleasures !" Order XXI. — Sport. — This is a very brief record, in spite of its comparatively minute sub-division. The people who make it their business to search fo^ game, large and small, constitute the chief iteirl in the first section. Then come the mino^ performers, such as puppet-showers and buffoons, not to mention the largest class of all, the conjurors. Exhibitors of trained animals, an,d the so-called snake-charmers, return some 10,500. Acrobats and tumblers are from the various gipsy hands, and are found all over the country in small parties, each professing kinship of some sort or other with the rest. There are the professional vrrestlers, too, many of whom are permanently on the establishment of the State, as in Baroda and Itjiyspre. Others, as in the Panjab and the South Deccan,' keep 'schools for. the training of champions, aiid much is the ingenuity exercised to keep and to discover the secret of some special wile invented by one bf these trainers for the discomfiture of the pupils of his rival at the next village festival of the neighbourhood. Lastly, there are the people ehtertaiued in the institutidflS for the exercise of Occidental athletic or outdoor games, of which cricket, at all events has toeen successfully acclimatised- in one section of the community. Billiards too' is not without its votaries, who seem to have taken to it to some purposes, as while this is being written, the current journals contain a notice that " in consequence of Mr, * * * (a Mlisalmai^) having won the All-England Amateur Spot-barfed Championship, telegraiis' bf congratulations are pouring into London from all parts of India.' ^ '' " ^ 11 .; ) ■■•'!,■ ;■' 72. Sport, &e. ^ 73, (a.) Trftii?.eci animals 73 (6.) Conjurors, buffoons, &q. 73 (c.) Acrobats, tuinble]'s,.&c. 43,919 12,448 36,453 ' 42,184 73 (d.) Athletic sports,' games, '&c. 6,176 Total, , Games, ^c. t ., 97,261 Total, iSport - 141,180 jn If recent criticism is to be trusted, we ought to interpolate here the word ^' even."— J. A. B. .111 74 (a.) Earthwork, &c. 74 (6.) ' General labour Total 1,1347,740 23,820,277 25,468,017 Order !XXII. — Ea/rth'worh and GerperqlLappur, —In the provincial series of returns this figures as Order XXIII. ;. but itv was found tbat the Order assigned for, the paste or complex occupations, whichit w^s, expected to find in the schedules, was not required, so the series in the Imperial return was increased by the sub-division of the former Order XXIII. into two, that under consideratioii an^ J;hat of the indefinite occupations. In the former we have one or two very distinctive occupa- tions, but the bulk is included under the general title of unskilled labour, of which, as before abseryed,jnuch-i8 simply agricultural ,v.- Amongst the special- brattchesdf earth- work we may mention the well-sinkers, a peculiar class, originating, it is said, in the ^PiStern part of the Deccan, but found all over central arid north-western India. The process by which they live does not involve the selection of, the site for their operations, as tbis has to be done under the sanction of spells or divination by more learned people. , But once the spot pointed out, this class not only excavate aud pass out iihe earth quicker than others, but they are experts, tpo, at setting the rings of pottery or other material, by which the soft earth is kept from falling in. The tank diggers and probably most of the railway labourers returned are of this class, as it is said that there is the same difference betw'een their power of work and that of the ordinary day labourer as there is between the efficiency of the latter clasp in England and that of the ;trained "navvy.''. The rice-poupder is here entered wherever it is a special occupation, as in Bombay, Bengal, and Madras. In Burma, where it provides for numbers of temporary immigrants from the last-named. Presidency, it is returned under the general head. Order XXIII. — Indejmlte and Disreputahle Ocicv/pations!- — Of these the former, as returned, is by far the more numerous, as it includes those who returned themselves simply as " dependent," without specifying the occupa- tion concerned, and also those who did hot return occupation at all, as amongst some of the wild tribes of Rajputana, Assam, and the Bombay States. Order, XXIV. — IndepeTtdsid of Oecupation.'—Oi these we have four sections. First, those who have property which makes it unnecessary for them to work at ariy occupa- tion for their bread, of whom only 193,291 are returned. Secondly, those subsisting on alms, of whom, first come the people enjoying scholarships , and so on, and, secondly, the army of mendicants, which, as before re- marked, must be taken with the class demandr ing alms as a religious obligation on the part of the giver, based on the sectarian or ascetic character of the applicant. Taken in this wa.y, the mendicant band amounts to about five and a quarter millions of persons. Inailother sub-division we find those supported by the State, of whom we have already spoken. First, the pensioners, numbering with their dependents 360,293; then the prisoners and inmates of asylums. Of the latter there are very few, 4,783 in all The number of prisoners is admittedly incomplete at 80,500, and many mua have been assigned to the occupation haijitualto them when not incarcerated. The next subject for consideration should rightly be the extent to which the above sub-divisions of non-agricultural occupation are respectively combined with some form of agriculture, but as the return bearing on this point does not include the whole of India, it will be better to take up first the wider su|)ject of the distribution of function between 76 (a.)' Indefinite ; - 75 (6.) Disreputable, &c 1,395,348 167,633 Total, Indefinite, &c. 1,562,981 76 (a.) Property 76 (6.) Alms Total, 76 77 (a.) Pension 77 (6.) Prisonersj, &c. - , Total, 77 , Total, Independent of Work 193,291 4,115,243 4,308,534 360,293 105,166 465,4.59 4,773,993 Town and Country. It has been already pointed out that, according to the last column of the general Broportional table we are now considering, the only Orders of occupation in which the share of the urban population is not greater Jhan tkatipf the rural, relativelyjo, their - ' ' 4 112 respective masses, are pasture, agriculture, and the sub-division connected witli village service. Next to ttese comes the category of Labour, which is semi- agricultural. The Potter is but slightly ahead, and then approach the other main vSlage occupations of the Tanner, the Carpenter, and the Oilman. The small group of the exhibitors of animals, &c. is the only one of the Orders which does not show an urban proportion of more than double that of the general ratio of urban population as a whole. If we go into a little more detail and consider the sub-Orders, it will be seen that the growers of special products bear a slightly higher proportion than the other agriculturists, indicating the admixture of sellers of such w^res, and also, probably, the growth of the latter on plots of land within municipal limits. ' In the next order, the supply of places of entertainment is very much more urban than the distribution of the personal services. The reason has been already given. Fuel, again, provides subsistence for relatively more in the town than lighting. Of vehicles and boats, only the last are built in nearly even proportions in town and country. Supplementary requirements' are naturally more frequent in the former, where tastes are Wider, and only the drum-maker, the tool-maker, and the purveyor of armlets and flower garlands and the like, reduce the general group- average. The Order of textiles varies much in detail, and silk and dress are appareiitly the items most favoured by an urban community. Cotton, jute, and wool are, from the character of their of)erativei3, more prevalent than the above in villages, though the proportion is still much in favour of the town; Iron we may class in the same category in this respect as pottery and wood-working, and it comes fairly between the two. The collection of forest pl-oduce is, of course, a rural industry, and cane-work is but a little less so. The commercial element, owing to the misnaming specified in an earlier part of this chapter, is relatively less of an urban pursuit than might have been expected, but the middleman rises in the scale considerably above the rest of the group. In the matter of transport, though even cartage is attracted to the town to an extent a good deal above the normal proportion, there is evidence in the return to show that both this and transport by boat are largely recruited from the village and fishing classes respectively. Amongst the professiohs, the priest and mendicant, the astrologer, the hail-averter, and our constantly recurring friend the drummer, are the only instances in which the urban proportion is not three times that of its total strength. This, of course, is only to be expected. In the village, where all is ordered by custom, and position is hereditary, there is no room for what Plato calls Timarchy, or the uprising of the ambitious.' But in the town, the professional has now full scope for the exercise of his faculties, — Kotaeque per oppida buccse Munera nunc' edunt, et Verso poUice vulgo Quemlibet occidunt populariter. He who formerly had to blow the trumpet for others can now blow his o#n. He heads deputations and gets put on the committee of agricultural exhibitions, and is even eligible Eor the jury which may perhaps convict his higher caste neighbour. The last sub-order but one, were it concerned with property alone, would, no doubt, show a'high superiority on the part of the town, but the preponderance of mendicants in the rural tracts drags the ratio down to a comparatively moderate figure. City distri- In Supplement of the above general ., comment, it will be interesting, perhaps, bution of to compare with the distribution of occupations of . the . province as a whole, that of occupation. ^^^ capital town, in the case of four or five of the chief urban centres of India. This is done in the following statement : — Table C. — Showing the DisxiNCTiyE Features in the Distribution of Occupation in Five Large Towns. IHDIA. ' Bombay akd SllTDH. BEsaxii. Madras. BUEMA. Haidkabad. Order. Enral. TJrban. Total; Bombay. Total. Calcutta^ Total. Madras. Total. BAngoou. Total. Haidra- b4d. I. AdminiKtration i-*ao B'S6 3-25 S'lO 0-98 3-S7 2-45 i-gs 1-36 4"*4 4-69 ■IVfl n. Defence 0-07 rrs 0-23 o'76 0-02 0-47 0-09 ■i'46 0-41 i-63 0-55 9-75 III. Other States' Service - 0-11 0-S3 0-04 0-03 - -, ,1 - O'Oi — o-o6 0-04 0-S6 IV. Pasture, &c. 1-S2 0-S2 1-37 O'S-I 0'64 o'ii 2-01 0-32 0-69 0-60 2-48 Q'sn V. Agriculture 64- S6 ■IT-IS 68-24 l6-9i 6S'19 n-79 58-81 3-62 63-44 rsi 44-88 3-S3 TI. Personal and Domestic Ser- S'SS 9'SS 2-65 rgs 3-61 «"o« 2-96 . 9-66 0-78 S-74 6-65 «4-«7 VII. Food, &c. - - 4-49- 10-Oj 7-BO' 6-os 4-96 7'si 6-90 ■I0-13 9-9S tS-g4 5-70 'r»7 113 Table C. — concluded. India. BOMBAT AND SiNDH. Bekqal. Madeas. Bukma. HaideaeXd. Order. Rural. Urban. Total. Bombay. Total. Calcutta. Total. Madras. Total. Rangoon. Total. Haidra- bad. VIII. Wght and Fuel, &o. 1-lS 2,V0 1-62 o'e$ I'OB ■/■Oi 1-26 i-Cfl 0-91 ^'43 0-80 y'oS IX. BuUdlng 0-36 rS! 0-80 i'SI 0'28 «•« i-oo a-37 0-64 o-gs O'SS i-is X. Vehicles and Vessels 0-04 O'SO 0-08 o'a6 0-06 O'il 0'05 0-3S 0'19 0-36 0-03 0'-l2 V. XI. S»pplementary 0-28 ■fss 9-65 ,, ^76 0-4S »-4« 0-28 3-66 0-43 ■fSS 0'28 i'OB XII. Textiles and Dress ,3-76 Wig 5-42 ■fS'SS 2-60 vm 4-71 6'7i 4'93 6-6y 6-28 3'l,i XIII. Metals, &o. 1-13 3'2g 1-61 B-6l! 1-13 3-0$ 1-38 if-aj 0-91 «•« 1-49 s'-f6 XIV. Earthenware and Glass - 0-81 o'gx 0-62 o'H. 0-67 o"4S 0'61 O'SO 0'60 o-eg 0-81 0-37 XV. Wood and Cane i-io rsg 1-74 s'm 1-17 JV« 1-78 S74 2-86 6'iiO 1-42 1-i3 ( XVI. . Prugs, Dyes, &o. 0-U 0-36 0-12 O'-IS 0-17 O'SO 0-12 o-ifi 0'15 0-27 0-17 ou/f XVII. Leather 1-09 r6a. , 1-28 ^•■fs 0-48 rsz 1-38 o-m 0'26 0-J7 1-37 o'gs XVIII. Commerce 1-25 s-is 2'12 ■n'aS 1-68 g-6s 0-97 5Sg 1-73 7'94 1-53 lf-76 XIX- Transport, &c. - i 1-06 i'ss , 2-02 a-66 1'27 ■lo-os 1-49 7-is 2-54 13-SS 0-73 J' Si XX. Professions 1-60 S'SS a-13 s-S 1-92 7-oi 1-96 -lo-oa 2-74 i-O-i 1-19 J'rjs XXI. Sport, &c. 0-os o'og 0'08 o'o6 0-03 o'os 0-08 o'oa 0-07 o'la 0-10 O'/i XXII. General Labour 8-77 , 9-7S 4-B4 4S'iS 12 '51 7'so 8-41 irsg 3-99 -IS' so 12-09 S-73 . XXIII. Indefinite, &c. 0-Bl o'Sg 0-17 yvo 0-21 iS'63 1-16 rss 0-OJ, O'Si 3- 83 ■rss 1 XXIV. Independent, &o. 1-48 3-4S 1-96 S'SS 1-13 S-6S 1-17 s'is 6-49 %-62 2-80 7-4* It is worth while, alsoi to note the two first columns, relating to the rural and urban distribution of the whole country, both as a standard by which to measure the special instances in the later portion of the table and in connection with the remarkti made on the general subject above. It can thus be seen at a glance that, speaking generally, the town differs from the country in its higher proportion of servants, food-suppliers, textile-operatives, traders, professionals, and the class that is inde- pendent of work, not to mention those engaged in administration and defence. It is also important to note how the introduction of the small town raises the proportion of agriculturists, for amongst the urban population, as a whole, the cultivator is still the best represented of any individual class. Passing on to the more special portion of the table, we may just_ touch upon the main points of difference between the five cities selected as most in contrast with the rural part of the Indian community. First, as to the agricultural element. Herein Bombay stands out high above the rest, since the municipal limits, being coterminous with those of the island, include a considerable area of " batty " or rice land. Similarly, this isolation from the neighbouring mainland and the general character of the cultivation on the latter, which is confined to one or two products, and no large surplus of those, seem to be the causes of the comparative paucity of the oil-pressing class and of those who bring in fuel and forage by headloads. The high ratio of the suppliers of food in Rangoon, and perhaps in Madras also, is apparently due to the number of the fishermen. The extension of textile industry m Bombay is well shown in the table under review, where the proportion stands at 15-82 per cent., as compared with 6-71 in Madras, which comes next to it. The latter, however has a' slight pre-eminence in metal-working. Rangoon, with its large teak yards stands first as regards dealings with wood and timber, and it is in the same position in respect to transport, since its shipping connection has of late years tended to outgrow the development of the permanent or shore population. In the matter of commerce or to put it more correctly, with respect to the proportionate number of nersons supported by that class of occupation, Bombay exceeds the rest considerably, with 11-28 per cent, against the 8-63 in Calcutta, but in transport the latter outshines it Madras stands first in its predominance of the professional class, with 10-08, against 7-bl in the City of Palaces, and only 5-16 in its western sister. The great influx of labour from Madras gives Rangoon the lead in this respect, but Madras itself is not far behind SO it is fair to presume that some of the contingent in the latter case are mallv deibitable to some more special head. There is a considerable number' coming under this class in Bombay, where the dock-works absorb, no doubt, all that the rotton mills can spare, but the relative paucity of the general labourer m Calcutta, tna-ther with the remarkable proportion of the entries grouped as mdenmte, which is 13-63 per cent, against 1-38 in the next highest case, leaves room for the suppositipn / 78388, P 114 that the tabulation, not the class jtself, is deficient.,, This assumption receives some confirmation from the equally remarliable proportion of the order relating to domestic service and so on, in which the head " Indoor Service " includes more than double the number found under it in the far more populous city of Bombay, so that it is probable that a good deal of clerical or general service has. been thus disposed of. The great diflference in some respects between the four cities of British India and Haidrabad, the capital of the dominions of His Highness the Mzam, will not have escaped attention. In the first place, there is the very large excess in the proportion of the administrative element, which is 14-12 per cent. On the total population, as compared with 4-98 in Madras, the highest rate in the selected cities, and 5'26 for the whole urban population of India. Defence, again, is 9*75 as against 1-75. Private service is 24-87 per cent., and in the other cities, as we have seen, the highest' ratio is that in Calcutta, where it stands at 15-08. Lastly, the independent classes in Haidrabad bear the proportion of 7-48 against 5-45 in Madras, and a general ratio of 3-45. Looking into details, it appears that pensions and mendicancy' are responsible for the greater portion of the Haidrabad figure. On the other hand, the trading, professional, labouring, and some of the commoner artisan classes are curiously deficient in Haidrabad. The weavers and carpenters, for instance, are in proportion fewer than in the surrounding rural tracts, instead of being in excess, as they are in the other cities. These divergencies are in accordance with what was said in a former chapter regarding the predominance of the ofiicial element in the constitution of the capital of a native chief. The main classes to be considered are his immediate entourage, his troops, and the couriers, with all their servants and dependents. Even in certain branc]ies of luxuries, recent events have shown that in these days of rapid communication encouragement is not always given to local talent, as it was of yore. The alternate columns of the table under review afibrd the means of estimating the difference between the distribution of occupations in the town in question and the total of the territory of which it is the capital. It is not worth while to comment on this in detail, as the figures mostly either explain themselves or can be explained from what has been said already. Occupations combined with Agricultbre. Information on this point has been collected., as observed in an early part of this chapter, for most of the largei' provinces, though in many cases the results are evidently incomplete. This is not altogether due to defect in the original record, but partly, also, to negligence in tabulation. It is clear that in every rural enumerator's block there are likely to be duplicate entries of all the main occupations'; namely, that of the occupation itself, and that of the same in combination with some form of cultivation or tenure of land. To save trouble in abstraction and subsequent addi tion, the latter was ignored by a lazy operator in the census centre, and amongst some 15,000 of such employes the proportion of the unjust is naturally high. This accounts, it appears, for the difference, in Table D opposite, in the proportion of the semi-agri- culturist in Berar, where the staff was chiefly of temporary hands, and Bombay, where the work was done by the more experienced agency of village accountants. In the case of Assam and the North-West Provinces the results show that trouble was in no ways spared to get out the full tale of this class of occupations, and here, too the supervision seems to have been eflBcient, and the proportions are the highest.in India. Comparison was made above, however, with Bombay, because the village system in Berar is practically identical with that of the western Presidency, and differs from that of the north and east. There are other instances in which the results are deficient, not because there is less addiction to these quasi- agricultural pursuits in the Provinces or States in question, but from defective tabulation. On the whole, we may accept the returns of Assam and the ISTorth-West as very nearly representing the actual facts, and Bombay as approaching accuracy in a less degree, whilst the rest are decidedly below the mark. The omission of Bengal, the largest Province, should be explained. The Census Superintendent of that heavy charge states that owing to one reason or another this return was so incorrect for a considerable portion of the Province that he thought it waste of space to include it in his series of tables.* He comments, however, in his Report on the details of about half the population, for which he considers the return is satisfactory, and, as the proportions he quotes are in general accordance with those for most of the rest of India, his assumption is * He seems,- however, to liiive pririud fi return of some port, Un it has not yet reached, the IinDerial Office. , .■ , ' 115 Table D, showing the Per.rcentage on each Order of the Persons returned as also partly Agricultural. BRITISH PEOVINCES. FEUDATORY STATES. Oedek. Madras. Bom- bay and Siiidh. N.-W. Vto- vinces and Oudh. Pa.njiLb. Central Pro- vinces. Berto. A.ssam. A]m(5r. IHaidra- I liad. I Bom- Baroda. bay Status. Pan.iflb Statwi. Central Pro- vince Stales. N.W. Pro- vince States. I 12. I. Administration II. Defence III. Ottler States' Ser- vice. IV. PasUu'e V. Agriculture* VI. Personal and other Service. VII. Food, &c. - VIII. Light and Fuel, IX. Buildings X. Vehicles and Ves- sels. XI, Supplementary XII. Textiles and Dress XIII. Metals, &c. XIV. Earthenware and Glass. XV. Wood and Cane XVI. Dyes, Gums, and Drugs, Ac. XVII. Leather, &o. XVTII. Commerce , XIX. Transport, &c. - XX. Professions XXI. Sport and Gtames XXII. General Labour - XXIII. Indefinite, &e. - XXIV. Independent Total 7-08 !5'67 i-IS 8-95 14-42 S-43 5-15 9-05 4-79 12-26 2'77 0-34 0-46 3'33 6-54 11-56 S-26 6-80 7-02 14-27 20-39 ■8-40 9-70 17-40 7-6.5 7-04 4-71 12-60 4-62 2-90 44-68 6-62 18-79 1-73 4-75 0-75 21-17 87-67 10-25 8-49 3-07 70 2-65 1-75 18-59 2-00 .-08 13-04 12-07 23-78 9-31 16-36 13-94 4-44 15-34 11-56 U-07 5-97 2-05 7-08 9-24 33-67 27,-33 12-24 21-86 31-79 28-45 19-35 10-68 14-97 18-65 10-06 18-28 20-20 7-75 12-08 14-42 6-18 2-42 1-03 7-96 0-02 9-17 1-57 1-66 3-7.7 10-28 7-03 7-58 7-04 17-99 4-54 7-74 2-71 5-66 3-97 0-34 1-21 2-ie 2,-76 1-64 2-72 0-42 1-31 10-28 3-78 0-37 1-01 1-37 2-82 2-64 6-47 4-92 1-97 76-50 26-01 42-90 51-79 14-66 29-53 16-90 0-22 38-02 4-39 5-87 20-41 21-1-3 9-40 2-10 7-40 7-08 6-38 10-46 18-47 20-85 0-09 9-01 2-06 3-03 0-19 0-72 0-13 6-21 3-25 7.-37 0-65 10-40, 16-47 10-24 ,2-60 2-63 0-64 1-80 0-71 2-57 6-68 3-74 2-09 0-83 13-36 12-11 27-03 5-18 7-06 0-95 9-46 1-86 4-69 2-14 6-48 2-10 0-93 0-47 1-47 1-67 4-27 0-20 1-16 1-76 3-77 0-49 1-07 1-96 9-12 3-71 6-62 13-06 13-48 6-78 14-36 1-23 0-S4 2-47 10-82 8-87 36-73 8-74 2-90 25-13 8-91 2-52 4-08 1-60 6-66 ' Order V. and the oocupatiDrs combined "vvith it are excindedr P 2 116 probably correct. Having regard to the varying degrees of accuracy adnaittedly to be found in the statistics on this matter, it is useless to review the return in detail. But there are a few interesting points that are brought out in "the figures as they stand. The classes that most largely combine some form of agriculture with their special occupation are, first, the village menials, such as watchmen, then the artizans we have grouped under the head of the normal village community ; the carpenter, for example, and blacksmith ; still more the potter and the oil-presser. The barber, too, is not an unfrequent landholder in some parts of India, and the tanner, the cart owner, and even the general labourer, eke out their income in this way. The weaver, another of the large rural industrieSj seems to be somewhat averse from this pluralisrh, and his own occupation, indeed, is hardly one that would fit him for strenuous outdoor work, so that unless he can get hold of a plot of land and let it out to others, it is difficult to see how he could make anything out of his possession. In fact, the proportion of weavers, as it is, must be largely recruited from the class who work, not in the more respectable fabrics -of cotton and silk, but in the humble hemp or flax, which is, as we have seen, the monopoly generally of the watchmen class, who are all semi- agricultural by tradition and predilection. Next we find a few professionals m the field, for instance, the proportion of the priestly class is high in .the following cases ; — Bombay, 17-26 per cent. ; Madras, 18-84 . Central Provinces, 22-33 ; North- West Provinces, 32-21 ; Panjab, 23-51, and Assam, 45-66. Even in Berar, it is considerably above the general proportion of the non-agricultural community. The religious orders, it is to be understood, are prohibited but to a small extent from worldly pursuits, and to a still smaller from the acquisition of property, in India. In Burma it is different, and the Buddhistic system is more severe in its restrictions. The Brahman, however, may do practically anything but hold the plough, and the very wealthiest are not debarred from receiving the alms to offer which is an act of religious merit. At places of pilgrimage, where the right to receive the offerings of the faithful is distributed according to hereditary position, a Brahman rolling in wealth will pursue, even to the High Court or Privy Council, a claim to the coppers of the potter or oilman — non olet. The lawyer, again, has his eye on land in the present day, and the fact that so shrewd a class should think it worth while to invest in this form of property is a testimony in favour of the current system of assessment and administration. In the following Provinces, various kinds of limbs of the law are returned as interested in the land: — Assam, where, however, they are comparatively scarce, 41-75; North- West Provinces, 15-16 : Bombay, 13-51, and Madras, 18-98. Lastly, we come to a question of considerable inportance, on which the evidence of the census as it stands is not at all re-assuring. This is the extent to which the land is passing into the hand of the money-lender. It must be understood that the subject is a widely spread one, on which but superficial evidence can be obtained from the data before us, but in nearly every province that has sent up fairly trustworthy returns, the proportion of quasi- agriculturists is higher amongst the money-lenders than any other class, save, perhaps, village menials. It may be that some of the persons included are really of the land- holding class, for, as has been already remarked, it is a very ordinary custom for a man, not of Islam, who has surplus cash in hand, to lend it on such security as he may think proper, whatever his special occupation ; and as such transactions would have been set forth in his income-tax return, it is probable that he thought it consistent to enter the same at the census. On the other hand, the wide extent of the high proportion, and the remarkable difference between it in British Provinces and in the two large Feudatory States, where, for the most part, the transfer of land to a creditor is unusual, if not unlawful, seem to indicate a real advance of the commercial class engaged in this particular branch of transactions on the land, as occupants, de jure or facto. The following are the instances in point : — Per-centage of Landholders, &e., amongst State. Per-centage of Landholders, &c., amongst Province. Total Total Non- Agricultural Population. Money-Lenders. Non-Agiicultural Population. Money-Leuders. Bombay 9-24 31-22 Haidrab4d 5-21 15-31 Madras 6-54 17-77 Baroda $-68 2-60 Central Provinces 5-.'>6 36-74 Bombay States 4-29 Berar 2-54 23-21 Central Province 10-82 13-48 Assam 38-02 67-65 States. K.-W. Provinces 18-28 46-57 Panjab 7-96 18-37 117 Tie excess in this special instance over the general propoi;tion of semi-agricul- turists in the proviilces is a very noteworthy feature in the return. Provincial KETtRNS. To conclude this chapter, a table is given on page 119 in which the relative strength of the 24 orders in the different Provinces and States is shown. The most convenient way to treat these details in the cursory examination which alone is necessary in this work will be to deal with each in its serial order, instead of taking the Province or State as the unit of comparison. The first Order will be seen to be high in British territory, according to the relatively complete equipment of the village staff. In Bombay, Berar, and Madras, the .proportion is high, but in Assam, Bengal, and Burma it falls. In Feudatory States, on the other hand, the proportions are usually high. In the case of Central India and the Madras group, Order III., should be added to this head, as it clearly belongs to it, but has been misplaced by a misinterpretation of the meaning of the headings. In the same way, Order II., Defence, has got mixed with the first in some of the States, owing to the adoption of the title of Service, unspecified. In British territory, this Order is mostly insignificant, save in Ajmer, where there is a large cantonment in a small province ; in the Panj^b and in Burma, where the number of trbops is always considerable. In the North- West Provinces the case is the reverse of that of Ajm^r, and the very considerable garrison is swamped by the enormous rural population. Pasture calls for few remarks. It is apparently most frequently returned in the Central Provinces and Madras of the Provinces, and Baroda, Haidrabad, and the Bombay States, of the rest of the country. Rajputana, too, shows a good proportion, whilst it is singularly weakly represented in the damper climes of tlie west coast, Assam, Bengal, and Burma. Agriculture has been sufficiently dealt with above, but attention should be again called to the fact that in the Madras States, Haidrabad, Ajmer, and Central India, the field-labourer has been very largely relegated, to the category of general labour, a fact sufficiently obvious from the return now under review. The very high proportion of agriculture in Assam, Ooorg, and the small States under Bengal and the Central Provinces may be noted. Order VI., dealing with service of various sorts, has been described already as unduly comprehensive. The proportion is high in the Panjab and in northern India generally, and is at its minimum in Burma. It is above the average, too, in some of the larger States, so presumably the proportion is dependent upon general service, as well as upon the special attentions of the village class of functionaries. As regards the supply of food and drinks, &c., we may note the high rank of Burma, where the fisher holds his own well, and in the Madras States, where he is likewise a great feature in the community, with the supplement, in this group, of the toddy-drawer. The proportion in Assam and Bombay is affected in the same way, the former by the fisher and the second by both. The provision of Light and Puel, &c. is remarkably evenly represented throughout the larger items, but the proportion falls in Coorg, Burma, and some .of the smaller groups of States. Building seems to be chiefly returned as a separate occupation in Madras, Bombay, the Panjab States, and Ajmer. The next Order, too, is too small to need much comment. In Assam and Burma the canoe digger and boatbuilder, respectively, are the main items, and Ajmer apparently owes its predominance to a colony of people on the railway. Supplementary occupations are too varied and too small to require notice here. Textile industries are proportionately high in the Panjab, the Central Provinces, and Haidrabad, and low in the more primitive communities of Assam, the smaller States of the Central belt of hills, and in 'Coorg. Nor do they flourish to any considerable extent in Bengal. Metals occupy a singularly even proportion throughout, except in Assam and Burma. It may be remembered that this group includes two or three of the main village industries, which accounts to some extent for the uniform distribution. The potter, too, is fairly distributed between a maximum of 1-83 per cent, in Rajputana and 0-26 in the North- West Province States. He is at a discount, too, in Coorg and Burma. P 3 118 The order connected witli working in Wood and Cane is a very yaried one as to distribution, in spite of its prevalence as a village industry. This is due to the inclusion of timber-dealing and work in cane and leaves. The former predominates in Burma, the latter in the coast States of Madras, and here the general ratio is highest. In Mysore there is a singular deficiency of carpenters. In the Central Province States, where huts are the usual form of dwelling, the absence of this artisan is intelligible. The small group of those working in connection with minor Forest produce and Dyes, &c. can be passed by. Then come the leather-workers, who seem to be in force more in the north of India than in the Deccan and south or east. In Assam and the small Central Hill States, indeed, the tanner and shoemaker is but a minute element in the community, and in Burma and Coorg he stands relatively little higher. Commerce, a mixed group, as has already been stated, is at its maximum in Ajmer and the Madras States. The former is essentially favourable to trade, from its position, and its proximity to the refuge of the most enterprising class of traders in India. But the secluded coast States of Madras seem hardly to compete in this respect, and the high proportion there is possibly due to some error in classification, as in the case of the' agriculturists. The census administration of the largest State in the group, Travancore, showed signs of originality in several respects, of which is this probably one. Under the head of Transport we have to consider that by water and that by road or rail. Bombay and Burma come first with respect to the former, and Rajputana, where the' railway is of comparatively recent introduction, shows a high proportion of road-carriers, and also, as we have said above, of porters and warehousemen. Professions, which to a large extent means priests and so on, are pre-eminent in Baroda, the Panjab, Burma, and Eajputana. Then comes Bombay, British and Feudatory. Except in the small States under the Central Provinces, they do not fall below one per cent., and on an average, they reach nearly double that proportion. Passing on to General Labour, we may note that this item is almost absent in Assam and the Panjab, and exaggerated, as we have described, in several of the States, if not in Bengal also, at the expense of the agriculturist. The indefinite group is recruited largely from Rajputana, where a good many entries were omitted, and from Madras, where there was a superfluity of entries of " dependent," without further qualification. In Haiderabad, too, this class is strongly represented, but in the absence of the local Report, explanation is not possible. Those who are supported by other than their own work vary greatly in distribution. On the whole, they are in greater force in the States than in the Provinces. Rajputana takes the lead, followed by Haiderabad and the Panjab. Ajmer, Central India, and Kashmer all show a high proportion. In most cases the ratio no doubt depends on the number of those supported by charity. In and near Ajmer, for instance, there are several places of pilgrimage, which form an outlet for almsgiving. In Rajputana the community is highly, orthodox, and the numerous Chiefs' courts are therefore much aifected by the medicant orders, whether devotees or others. The Panjab has to support two religions, each with its separate brain of claimants for the charity of the faithful. Presumably, the corresponding class in the Madras States has been endowed with land or some other means of livelihood, as the proportion shown is remarkably low, considering the number reported to be dependent on the royal bounty in the "Land of Charity " along the Malabar coast. With this comparatively brief summary of the diverse conditions of occupation in India, in its various sub-divisions, this review must conclude. The reports of the Provincial and State Superintendents of the Census contain much detail that wiU repay perusal, but which it is impracticable to incorporate in a general work such as the present. 119 Table E. — Showing the Per-centage of the several Orders in each Province, &c. PbovinCk on Statk, &c, □ O I a 'a ■ < 1 u .a 1 1 ■i 1 i t-t (D . '^ I S3 ^^ '^ s 0) i n p2 1 B > '> 1) d 1 A. ft ■1) c; 1 ■'. y z l/I. II. III. IV. ■ V. '. VI. vn. VIII. IX. x. XI. XII. ], Ajmer 1-28 I '06 0-41 I "21 48-40 6-37 2-36 I-07 0-77 I -16 0-56 5-66 2. Assam 0-67 0-17 O'OI 0-28 76-91 1-62 6-87 1-24 0-27 0-20 0-21 1-37 2. Bengal 0-98 0'02 — 0-64 63-19 3-5i 4-96 I -06 0-28 0-06 0-43 2-60 4. Bombay and Sindh 3-2.5 0-23 0-04 1-37 58-24 2-55 7-56 1-52 0-80 o-o8 0-65 5-42 5. Berar 2-98 a- 07 — 1 '21 68-55 2-50 2-90 1-54 0-62 o-oi 0-35 3-87 6. Burma i-36 o'4i — 0-69 63-44 0-78 9-93 0-91 0-64 0-19 0-43 4-93 7. Central Provinces r82 0- II — -•47 64-04 2-44 4-10 2-00 0-40 o-o5 0-47 6-74 8. Coorg 2-20 O'OI - 0-72 72-00 2-39 5-62 o-3o 0-38 0-02 0-19 i-jS 9. N.-W. Provinces and Oudh 1-80 o'lS 0-02 0-83 60-81 S-41 4-90 I -64 0-32 o-o3 0-66 4-f.7 10. Panj^b - - i ^ i 1-79 0-S2 0-08 I'29 55-65 6-73 5-16 I -04 0-59- 0-04 0-24 7-56 11. Madras 2-45 0-09 ■ — 2-OI 58-81 -"•-gS 5-90 1-26 1-00 o-o5 0-28 4-71 12. Haiderabad 4-69 o-SS 0-04 2*46 44-88 5-65 6-70 0-80 0-53 o-o3 0-28, 6-28 13. Baroda ... 356 0-82 0-03 2-27 57-76 2-98 4-12 I -45 0-58 0-02 0-28 5-14 14. Mysore 4-32 0-45 — 0-47 66-61 2-21 1-27 0-47 0-62 0-02 0-20 2-94 15. Kaslimer 2-25 0-42 0-03 i-i8 68-12 3-33 2-38 0-59 0-12 — 0-16 5-73 16. Rajputana 2-67 I>23 0-21 2-34 52-69 5-26 2-66 0-93 0-49 — 0-24 3-o3 17. Central India 0-74 0" 17 3 -.38 0-94 48-12 5 10 2-31 i-o5 0-32 — 0-34 3-21 18. Bombay States 3 •.'55 0-2S O'll 7. -63 57-33 2-35 4-02 1-53 0-68 0-02 0-27 5-72 19. Madras States > 0-01 0*04 1^90 o'3o 43-27 2-99 8-54 1-26 0-64 o-o3 0-11 3-45 20. Central Province States 1-85 — — 2'72 70-62 '■47 3-50 o-8i 0-07 o-oi 0-37 7-68 21. Bengal States 1-18 O'OI — 1-87 69-19 2 -20 2-59 0-55 0-11 0-02 0-13 2-97 22. N.-W. Province States 1-94 0-76 0-Q2 o'4i 69-94 4-91 4-26 1-09 0-30 — 0-29 4-5i 23. Panjab States 1-97 0-46 0-05 i-3o 64-55 4-56 4-25 0-83 0-87 0-04 0-15 5-48 INDIA 195 0-23 018 1-27 59-79 3- 91 607 1-23 0-50 005 0-40 4-39 P 4 120 ( Table E. — concluded. Province ok State, &c. 1. Ajmer 2. Assam 3. Bengal 4. Bombay and Smdh 5. Berar 6. Burnla ?. Central Provinces 8. Coorg 9. N.-W. Provinces and Oudh 10. Panjiib 11. Madras 12. Hyderabad ■ 13. Baroda 14. Mysore 15. Kashmer 16. Eajputana 17. Central India 18. Bombay States 19. Madras States 20. Central Province States 21. Bengal States 22. N.-W. Province States 23. Pa'njab States INDIA : ( - XIII. 1-47 0-75 1-13 1-51 1-32 0'91 1-57 1-59 1-37 1-64 1-38 1-49 a-51 1'49 1-00 1-22 1'27 1-57 1-71 1-2.5 1-41 1-40 ■l-.'il 1-33 d Caste. Afi 8' ctiyvcufjiovcos a'Koue«v irfpt t% 'luhKr^g ' Koii yap aVfOTOtTto Vo-TJ . . . o» 8e Koii iBovreg y.eprj T*i/a et^ov, TO. he irT^eiio "Keywtriv e| a';fo% . . . hioTrep oo'Se to. motm Trepi rCov auToiv e^ayye^.Xouj(;; ,■■■ •'.•■tr>- ■■-■■ ■- •: ,' •• ■ - -- ' ,;.,•- / 78388. ' K 130 audi matters. This review has to be limited to the statistical side of the questions involved, that is, the relative strength and the local distribution of the numerous component elements of this heterogeneous community. There renaain untouched the, more interesting topics of customs and folk-lore, information on which forms the most valuable contribution that India can make to the study of ethnology in the present day. For, as has been pointed out of late more than once, speculations on these subjects have been too often based on the comparatively unverified accounts of customs observed amongst the lowest type of savage, whilst in India there lie ready to hand abundant instances of all sorts of custom, barbarous and otherwise, in. every stage of transition or development. During, the last 20 years some advantage has been taken of these stores, as can be judged from the works of Mr. Campbell in Bombay, Mr. Eisley in Bengal, Messrs. Orooke and Kitts in the North- West Provinces, and the ethnological chapters, still unrivalled, in Mr. Ibbetson's work on the Panjdb. In the light that it has been possible to throw on the general ethnographic position in India by the foregoing remarks, there can now be considered the distribution of the population according to, first, language ; secondly, religion ; and lastly, race, caste, or other social subdivision. Scope of the inquiry. General results. A.. — Tee Distribxition of the Population accouding to Mother Tongue. "Ea-Ti he xa'KKa. i'dvea 'IvSuiv, Koii orjK b[JLo Influence of an exclusive literary class, Results of BuddMsm on language distribution. pMlologist, as it ought to beco^efitrbat of tke ethnologistj^for lieretwe fiird.^ ianguagfl in every stage of its developmeiilt ; or, to ■ put it otherwise, forilis df ^s^eeolt! dre '-here current which, appertain to nearly every one of; the^ main classes recognised an philology. Its geographical isolation, almost as much as its political history, have contributed to this, as can be judged from the general sketch of the ethnography of the country above given. -Especially in the case of language have the ■ influences to which it has been subjected been peculiarly local and restrictive. A mountain range, a belt of dense, dank forest, the deep- gorges of the Himalaya, so completely separate the respective inhabitants of their -.flanks that communities which have split off" -from the same tribe but three or four generations ago are often unable to understand each other's tongue. Politically, again, the spread of the more civilised along the plains and great river basins has driven a considerable portion of the oldeivt population to the shelter of the hills, which thus stand out like patches of cover in the midst of cultivation, afibrding a place of refuge to all the wild animals that have been expelled from the rest of the country. Nor was the immigration itself i more favourable to the maintenance of the integrity of the- languages thus introduced) into the country. Setting on one side the Skythians, of whose tongue praicfcically no trace remains, the only movement of importance, from our present standpoint, is that of the Arya. .Now the progress of this community was not that of a united race or nation, but of various collections of clans which were disintegrated with great rapidity into small independent bodies. The growth of ^he community, therefore, numerically, seems to have been in advance of that of their civilisation, and by the latter alone can a language be spread uniformly through the mass of the population. Then, again, the intercourse with the darker race must have had its effect on the tongue of the immigrant, and its influence is likely to have been far greater than that which is usually assigned to it. But the establishment of a number of separate States and free admixture with the daughters of the land were not the only influences' that helped to break up the unity of the language brought with them by the Arya. The firm establishment of that race in the plains was followed by a highly peculiar ecclesiastical development, which will be noticed in its place in the next section of this work. But as it had a direct influence on the linguistic tendencies of the commiunity, its general features must be briefly described here, in anticipation of a more complete account. It began, then, with the elevation of the family priest into the member of a hierarchy, so that, from being the agent of the tribal or family patriarch, in the ceremonial of sacrifice, he became the controller of all social as -well as religious ritual. The sacred formulae became his monopoly instead of the special appanage of the chief, and it is not difficult to conceive how this change weakened the hold of the original language amongst the other classes. For the farther the latter spread over the country from the primary settlements, the more exaggerated was the value the priest was able to place on the exact knowledge of the sacred words, and the closer was the restriction of that knowledge to within the hieratic class that he was able to impose. There was accordingly no linguistic standard available to the masses, and, in the end, it was made penal for any but the upper classes to even listen to the recitation of the texts. A survival of this triumph is found in rthe present day, when in some parts of India it is the practice of the impure castes of the Brahmanic system, who are the direct descendants of the helots mentioned above, to get an ecclesiastical sanction for their unions by performing their wedding within sight of, though at a distance from, the corresp^ondi'ng ceremony amongst the orthodox, who are being tied tbgether with full rites.' Until comparatively modern times, the prohibition in question ainounted practically to the denial of instruction of any sort to the masses, since the greater part of the learning here, as in so many other countries in the corresponding stage of their development, consisted solely of expansions of various forms and uses of the ritualistic texts, known only to the ecclesiastical body. A temporary break in the continuity of this system was caused by the rise of Buddhism, a sect which repudiated the monopoly of the key of salvation by the priestly class, and was the means of difiiising a certain amoupt of , instructiqn amongst the middle classes also. There were, about the time of this reformation, two main off'shoots of the original tongue of the Arya current in India, the Mag9,dhi and tl^e Sauraseni. It is an open question, perhaps, whether these divisions represented real distinctions in the vernacular, or were only the creations of the grammarians of a later period. At all events, the new faith was preached in the JMagadhi dialect of what i^ known as Prakrit, which was prevalent, , it is said, in;the,.e^st and north .of thf rallying point of Brahmanic q;vilisation,tha,t is, the mlji^le tract^ of'/the GangeS;.9,n4 Jamna 133 Doab. This dialect was tJnjs carried by Buddhistic missionaries far into the Himalaya and Hindu-Elisli, and it is possible that it may have been the. foundation of those now current in the outlying portions of Kashmer and in Afghanistan. In the latter, however, the original tongue of the Pakt^i of the Greek writers, the modem Pakhtwan, of Pathan, as they are termed in India, has been overlaid to a great extent by a veneer of Persian received from the west, so that its construction alone remains to indicate its Arya origin. In another direction, it penetrated into what is now Nipdl, where the religion it served took so strong a hold of the higher valleys that in later years, when the Brahmans, driven out of the plains by the Musalman invasions, fled into this part of the country, they were able to make an impression only on the lower valley. Here, however, they found consola,tion of a worldly character, and started a colony of mixed blood, which became a power in the land, and in the whirligig of time managed to incorporate the ruling race into the, orthodox pale, as being of the warrior caste of that system. The language introduced by them, with the invariable local modifications by the masses, became in time the court dialect, and obtained possession of the whole of the lower portion of the State. It has already been stated that a somewhat similar process of incorporation was Influence on adopted in the southern part of India, though here, as there does not seem to have been ^outhem any mixture of blood, at least at the time in question, the local dialect assimilated only the language] vocabulary required in ritual and in social ceremonies. It is not improbable that the intercourse with the south did not grow reaily close until the wave of Buddhistic monasticism spread over the peninsula, and crossed the straits into Ceylon. If this were the case, it would account for the peaceful nature of the conversion of the Tamil races, for the expulsion of Buddhism from India seems to have been a merely sectarian measure, and owed its completeness to the superior social attractions held out to the masses by the general lowering of the Brahmanical tenets to nearer the pre-Aryan level. The structure of the language remained tinchanged, and the vocabulary received useful additions, as was the case in Burma, to' which country the Buddhist missionaries proceeded after their success in Ceylon. The Hill tribes of the south followed the same course as their compatriots of the Central Belt, with whom they are physiologi- cally and otherwise inextricably mixed. The subsequent political changes that have occurred in India consist of either the Later short and sharp impact of alien races from the more temperate regions of Asia, or foreign movements of sections of the people of the country itself, neither leaving any marked imprint on the language of the community as a whole. We have, again, the occupation of the country by foreigners from a distance, who, whether they share the honour of philological Aryanism or not, are still farther from the parent stock than even the denizens of India at the present day, and come in such small numbers that even if they had the intention of permanent colonisation of India, instead of the reverse, they could only have an insignificant effect on the language of the places they thus favoured. On the invasions by land a short comment is all that is required in this section. Musalman The Musalman dynasties imported with them a foreign element that settled in the influence, country and administered its resources. They thus set their mark on the vocabulary, but left the rest of the language untouched. The dialect known as Urdii, from the Turki name for the citadel, or chief's camp, consists of the vernacular of the tract round the seat of government, interlarded with large numbers of Persian words, or of Arabic words received through the Persian.* There has been no introduction of the Persian construction, nor are the Arabic terms inflected according to their own rules, but they have to conform to the grammatic system of their host. Of Tiirki there is but a very slight trace in the vocabulary and none elsewhere in the language. But some of the conquerors sought to propagate their creed amongst the people under their rule, and the results on the language of the latter is curiously varied. In the north, where the conversion has been on a large scale, the vernacular has sufi'ered no change. For instance, in Baltistan or Little Thibet, where the population is Musalmdn almost to a man, the language is the same as that of their Buddhistic neighbours in * The account given by Bernier, of Dehli and Agra, which cities he describes as moving almost en masse when the Emperor shifted his quarters for the summer, has been quoted on page 46.. US 134 Ladakh, and only tlie special terms required by the new faith are introduced, and these are modified to suit the local pronunciation. In Kashmer, where again the masses are Musalman, it is the same with the Aryan vernacular. No change, too, is noticed to follow conversion in the Panjab. In its neighbour, Sindh, and in the Laccadiv Islands and with the Mappila of Malabdr, it is the same, though for corre- spondence the Arabic character has been adopted with a few modifications. There are, similarly, certain well-known mercantile bodies in the west of India who are all Musalman, but though they study Arabic for religious purposes, they have remained staunch to their native Gujardthi or Kachhi in all besides. So, too, the converted cultivators of GujarAth are undistinguishable in speech from their Hindu neighbours, and the colonies of probably Arab blood that have settled there, have adopted the same vernacular. On the other hand, there is all over India a numerous class of Musalman converts, especially in the towns, comprising artisans, domestic servants, and those who largely swell the bodies of the police and the native army, who have not only abandoned their original caste titles, and, as already stated, have affiliated themselves to foreign tribes, but regard the dialect of Hindi known as Urdii as the peculiar appanage of their faith, and adopt it, accordingly, with a strange and varied garnish of the local vernacular. Sikh and Maratha influences. As to the movements amongst the inhabitants of India itself, those of the Sikhs and the Marathas may be briefly noticed. One of the results of the success of the Sikhs in emancipating themselves from Brahmanic orthodoxy was the erection of Panjabi to the position of a separate language. This was an accident, like the estab- lishment of Vraj or the Dehli Hindi as the court language of the Moghals, with this difference, that the latter had already an established position, whilst the former cannot be even now said to Lave any recognised standard. Then, again, as the Panjab tongue was so nearly related to its eastern rival, there was no occasion for an aggressive propaganda on the part of the new State across the Jamna, and the Lov^er Indus valley did not tempt occupation. In the case of the Marathas, too, no attempt was necessary to extend their language. The object of their expeditions was mainly gain, where it was not to obtain possession of territory already inhabited by their own race under foreign domination. The invader took what was to hand,' and left a sufficient establishment of his own adherents to ensure the due realisation of future benefits, when the time came. In the meantime he was called to distant duties, and when ousted in his turn from his new nest, naturally nothing would be found of a permanent character in his arrangements, still less in the effect of his occupation on the language of the conquered tract. For instance, in Orissa there is no trace cf Marathi in the language, though the Maratha domination lasted over 50 years. In Tanjore, the Deccani is a myth save in the precincts of the late court. Influence of caste on the Indian languages. This summary of what may be called the dynamics of Indian philology, can be fitly brought to a close with a few remarks on the results of the influences above described in the present day. The particularism that has been noted as the leading tendency pervading the atmosphere of the country, has prevented the evolution of anything in the nature of a Ungua-franca, current throughout the continent. "What is termed Hindustani is generally the addition of a few conventional terminations to the local vocabulary, with the introduction of some Persian words. The language of the ruling race is known as a working means of communication to a few residents of the larger towns, and these belong to a very restricted class. The vernacular is apt to change, not only territorially, as it does from county to county in England, but by classes also. Then, again, the gap between the literate and the rest is so great that it can hardly be appreciated from the mere repetition of the fact that only six per cent, of the population is not wholly illiterate. It will be seen in a later chapter of this work that a very high proportion of this small number is found in three or four groups numerically insignificant, compared to the predominating interests of the country. Thus, it is not only the want of a standard in language that is felt, but the absence of root or depth. The expanding needs of the tongues in modern times are supplied not from the living language, but by reference to a classic, whose scope was for ages narrowed down to a few academic uses by a class desirous above all things to prevent its vulgarisation in the mouths of laymen. The diction of the literate bifurcates, accordingly, on all important points from that of the masses, on which it 135 lies like an artificial stratum, stifling the natural capacities of the " voice of the people." " Et sua mortifera est facundia.'' On this point the -words of the late Mr. Lowell are well worthy of consideration : — " It is only from its roots in the living generation of men that a language can be reinforced with fresh vigour for its seed. What may be called a literate dialect grovrs ever more and more pedantic and foreign till it becomes at last as unfitting a vehicle for living thought as monkish Latin. That we should all be made to talk like books is the danger with which we are threatened by the universal schoolmaster, who does his best to enslave the minds and memories of his victims to what he esteems the best models of composition ; that is to say, to the writers whose style Is faultily correct and has no blood-warmth in it. No language that has faded into dictiou, none that cannot suck up the feeding juices secreted for it in the rich mother earth of common folk, can bring forth a sound and lusty book. . . . Thei-e is death in the dictionary, and where language is too strictly limited by convention, the ground for expression to grow in is limited also, and we get a "potted" literature, Chinese dwarfs instead of healthy trees. . . . The schoolmaster has been busy starching our language and smoothing it flat with the mangle of supposed authority." So far as to the languages of India proper. The description of the remaining Mongoloid linguistic divisions will occupy but a short space. In the Himalaya and the high influences, lands that they support to the north, the Buddhistic propaganda secured a certain stability of language, by the medium of monastic institutions, but the basis of the tongues were unaltered. So, too, in Burma, where the same system, through its teaching agency, has given a power to the language of the majority that seems to enable it to absorb gradually most of the local distinctions of dialect prevalent amongst the minor tribes. The gap between the Burmese races and those of the central Himalaya is difficult to fill from the material to be found in the Delta, but some connection no doubt there was in times long past. The general course of the population of Assam has been already sufficiently described, and the tribal languages thus introduced seem to have remained undisturbed in the hills first occupied. On the descent of a community to the valley, however, freedom of communication seems to have been fatal. The Kdchh, for instance, has given place to Bengdli ; the Ah6m substituted for their own language that of their conquest, the Ohutia, and now the few relics of the former language can hardly be deciphered, even by the tribal priests, and are used only on great occasions, such as for rain in time of drought. . The Ohutia, in turn, have fallen before the Assamese of the valley, and in their case, too, the original tongue is known only to a few priests and exorcists. But in the Hill tracts agglutination is left to work out its own reformation. Reviewing the whole position from a philological standpoint, it will be seen that Linguistic India is hemmed in on the north and east by forms of speech of what is known as classification, the Tonic class ; in some parts, too, in the agglutinative stage. The central belt of hills has been left to forest tribes using an atonic form, rudely agglutinative ; and the south of the peninsula is almost entirely Dravidian in language, that is, the agglutination has reached a stage in which it is scarcely distinguishable from the Inflectional or Synthetic class. This last forms the basis of the language of the whole of the northern plains and the base of the Himalaya, with the Gangetio delta, the whole course of the Indus and the upper coasts of the peninsula, both east and west. It stretches, too, beyond the British frontier towards the west. The influence of the displaced Dravidian, however, is traceable throughout in the degradation of both construction and vocabulary from the Arya original, but in the highest stage of the latter, that is, in the region round its traditional centre, it is the least inflectional, or the most analytic, of all this class, approaching the western languages of Europe in its tendency in this direction. It meets in this stage the lagging Hill dialects above- mentioned ; and what with the opening out of the country by roads and railways, the administration of the forests on the modern economic system, and the reluctant advance of the village schoolmaster, even the wildest of those tribes are being brought within touch of the outside world. Means of livelihood are being disclosed to them of which a few years ago they never dreamed, and each step forward is accompanied by the acquisition of something fresh in the way of vocabulary, even if the adventurer does not pick up an entirely new dialect, with which the philologist of the future will have to wrestle, even as his predecessors strove with the mongrels of their day. R 4 136 The general classificatiQn of the langwage retmrn is as shown in the following table Table A. — Showing the GtEneral Linguistic Distribution of 10,000 of the Population. Language and Class. Number per 10,000. Language and Class. Number per 10,000. A. Indiq-Aryan Hindi Bengali Maratlii Panjabi Gujarathi "Criya TJrdii Sindhi - Western Pahari Assamese Central Pahari Marwadi Kachhi Halabi Kashmeri Goanese, &c. Eastern Pahari Minor Languages B. Dravidian Telugu TamU Kanarese Malayalani Gdnd Tiilu Oraon Kandh ■Kodagu Mal-Pahadia Brahui Minor Languages C. Eolariau Santhal Miinda and K61 Korwa and Ktir ■ Bhil Sawaria Kliarria Baiga, Bhinja, &c. Gadaba D. Gripsy Dialects £. Ehasi Dialects F. Thibeto-Bunnau Burmese Arakanese Kachari ITipali dialects Garo Tippera Khyin , Naga dialects Mikir Mech - ' 7,460 3,269 1,578 721 676 405 344 140 99 58 55 44 44 17 6 1 1 1 2,021 769 581 371 207 53 19 14 12 2 1 1 113 65 25 7 6 4 3 2 1 15 278 212 14 8 7 6 5 5 4 4 4 P. Thibeto-Burman — cont. Kathe (Manipuri) Miri-Abor Kliki Laliing Llishai Thibetan Minor Languages Or. Mdn-Annam M6ii {Talaing') H. Shdn (Tdi| Shah (iinspiecijisdy, J. Malayan ;^ X. Siuitic Kar^n Chinese L. Japanese M. Erinic-Aryan Pakhtii Baldch Persian N. Semitic Arabic 0. TurAnic ( tjgrian) P. European-Aryan English Q. Basque Ha mitic Not distinguishable or not returned - , 7 ■7 27 26 1 51 41 9 1 137 • .^^^^^^.o^ all come the tongues which are traceable to an oridu amonffsb the Aryan aa.i unmipants that have been termed the Arya, which are returned by about three^ ^'-^^''Ji^" .r^.Mi the population, gecondly, in numerical sequence, though longo iniervallo, «'°"P" comes the Southern or Dravidian group, in which are included the languages of the Hill tribes of the central ranges which form the northern frontier of the Deccan on the eastern side Ihey are almost isolated geographically from their comrades, and much_intermixed with the next class. The northern but entirely separate language. Brahui since it has been admitted to be morphologicaUy of the south, ff\ ?Tv?^^f included, though with considerable diffidence, ^nhalese, too, and its offshoot Mahl, or the tongue returned by the inhabitants of the lonely little island of Mmikoi, midway between the Laccadiv and the Maldiv groups, have been similarlv treated, swelling tne total to about one-fifth of the population. The small remains of an older type of language, restricted to the tribes of the hill tracts of Western Kolariaa Bengal and Central India, with a branch or two running west and south has been g'o"P- denominated Kolarian, though there are no doubt good objections to that name. It has been adopted, however, by so many distinguished writers on Indian philology and ethnology that though it may be wrong it is quite intelligible. The whole group bears a proportion of but a trifle over 1 per cent, of the total, and contains onlv two items of any considerable prevalence. There are two groups statistically insignificant, but having special interest of their Khi», and own m other ways. The first is the tongue of the inhabitants of the Khasia and the gipsy Jaintia Hills, between the two main valleys of Assam. It has not been affiliated to ^i^^^^ts. any of the surrounding languages, whether of the Aryan or the Tonic families. Till recently it was unwritten, but now, owing to the study and enterprise of the Welsh missionaries, the Eoman character has been generally adopted. The whole community, however, with its three dialects, only numbers about 178,000 souls. The second of these groups is that of the many tongues spoken by the wandering or Gipsy* tribes of the plains of India, and numbers just over 400,000. It is out of the question to distribute these langaages amoi;igst those having fixed dialects, as their character changes with the locality most favoured by the tribe using them, and, whilst retaining a backbone peculiar to itself, freely assimilates the local voca,bulary and pronunciation. The most prevalent of these dialects is that of the Brinjaras or Lambani, the carriers of Upper and Central India, which is based on a sub-Himalayan Hindi vernacular. The tribe, however, is found as far south as the Madras table-land, and it is not improbable that the Lambani of the Deccan qould hardly make himself understood by the corresponding caste farther north. Again, the earth-workers called Od or Waddar, carry a language of their own from Peshawar to the sea, using a vocabulary less and less Dravidian as the tribe frequents tracts farther away from the East Deccan, from whence it probably originated. More difficult still, as regards classification, are the dialects used by the less reputable tribes of wanderers such as the nominal Hindi of the thieving and mat-weaving castes of Hindustan, and the Telugu and Marathi of the acrobats and pickpockets of the Deccan. All these can doubtless be generally divided into degraded forms of either Hindi or Telugu ; but in doing so we have to disregard the local characteristics just mentioned, so they have all been taken under a heading ,of their own, namely, Gipsy dialects. We pass now into the Tonic zone of language, which comprises in all its branches Thibeto- just over 3 per cent, of the population. As in the case of the regions of India Burman linguistically Aryan, this proportion would be higher if the whole of the races groups, using language of this formation had been brought under enumeration. But a large tract on the borders of Burma and Assam had to be omitted from the census, owing partly to the unsettled state of the tribes just included within our territory, and partly to the difficulty of getting competent enumerators for a population not only entirely illiterate, but using a language in most parts which has never been reduced to writing. In tabulating the information obtained, it has been found convenient to group this otherwise unwieldly class by geographical position, beginning with Thibet, and working eastwards through Assam down into Burma, where the largest of the com- ponent units are found. Here the main group, called in the return the Thibeto- ' Burman, is touched by three smaller groups of a kindred class, though differing * The term Gipsy is hei'e used conventionally, as the equivalent of vagrant. The only class to which it now applies in its European signification is that mentioned at the end of this paragraph, the subdivisions of which are very numerous. / 78888. g jm enough, to be separabely shown. These have been already mentioned as the Tai or Minor Tonic Shan, the M6n, and the Karen. The two first appear in the tables • to a- very small groups. extent compared to the total number of the races by- whom they are spoken. The Tai group, for instance, includes a small colony or two in. East lAssam, and a few more representatives in Burma, but the bulk of the Shans are found either in the border States, where language was not returned, or in the neighhouring kingdom of ''Siam. The Mon language, too, is returned only in Lower Burma, and in an isolated' traot occupied by the Palaung on the frontier of Upper. Burma ; but the greater portion of the Mon race lies in Anam and Cambodia, or Khm6r. The Earens are dhieiy confined to Lower Burma, and their language has been cla,ssed with the Chinese, to which, according to the best authorities that have been consulted, it can be"< most safely affiliated. The Japanese language is . hardly represenlied. The Malay class is chiefly interesting, from an Indian point of view, on account of the curious group of sea-gipsies called Salon, inhabiting the.Mergui Archipelago in the south of Tenasserim. The Nikobari was returned by one individual only and as its correct' allocation has been yet satisfactorily settled, its single representative has. been allowed to bring up the rear of the Burman group, though possibly he would be more congenially mated with the Malays. Other groups On the same grounds as those on which Chinese has been grouped with Karen, of language, though the bulk of those who use it are foreign to India, this survey may include the Branic section of the Arya language, which is represented by a considera,ble number of border tribes on the north-west; and the Semitic, because Aden shows the majority of its small population to be speakers of Arabic. The Skythic or Turanic group is a very small one in the return, and the Hamitic, like Arabic, is almost confined to Aden, where the settlers or sojourners from the opposite coast are relatively numerous. As for the European element, all that need be said is thai the whole tale is about 24,6,0OD persopis, of whom 238,000.; return English. To conclude the list, one person returned Basque as his parent tongue. The statistics ThO' languages have now to be reviewed in -detail. It is as well to begin within of language. ^Yie largest group, which is also the most advanced, namely, the Indie- Aryan, or the Inflectional or Synthetic class. This- has been subdivided into three geographical sections, the n'orthe!m, the western, and the eastern. The ' first is by far the largest, as it includes Hindi, which pr4^ dominates throughout Hindustan and Bihar, two of the most thickfy peopled tracts in India. This title is admittedly a comprehensive one, and incltides all the varieties of Aryan speech between the Jamna and Rajmahdl. It has thus absorbed such distinctions as Priij, Baiswdri, and BTiajptUri, as wisll ''as Mhithili and other Bihar varieties-^rof Hindi. Hindi. The same is the case in the south and west of Hindustan, where BimdM, BdgMli, Bdngadi, and NimdM were returned by but a small fraction of the 'persons known to use a local variety of the standard language. On the other) hand, Ldria, the last form of Hindi found in the south-east, was returned pretty, completely. A want of detail of this sort is often attributable to the absence of literature in those tongues, a fact which led to the enumerator's hesitation to ; give, them official recog- nition in a formal document, such as he considered the schedule to be. It is noticeable, . too, that most of the returns of dialect were made by persons n©t belonging to the place where they were enumerated, so the census agents^ having foreigners to deal with, thought it most prudent to accept the term dictated to them rather than* betray their ignorance by contraverting its use. Then, again, as far as Hindustan proper is concerned, that is, the present North-West Provinces and Oudh, the language of the plains is officially " Hindustd,ni," so it is , not likely, that any attempt would be made through the medium of the census to contravene this authoritative decision.* On all these considerations, therefore, all such .sporadic entries have been absorbed into the general title, lest the separate recognition of such small and scattered numbers should give a misleading notion of, the actual prevalence Language. Population refumi'ng it. Hindi Western Pahari Central* ... Eastern Panjabi Kashmeri - - - Hindu-Kiish dialects Total, Northern Group 85,676,373 1,S23;249 1,163,233 24,262 17,724,610 29,276 17 106,130,020 * Garhwali, 647,739 ; Kumaoni, 505,494. * It is a question, too, whether the distinctions are recognised at all, otherwise than by grammarians. 189 of the dialects in question. It should be pointed out, too, that west of the Jamna, the boundary between Hindi, Panjabi, and Marwd,di is scarcely distinguishable ; and similarly, on the east, the tongue of North Bihar becomes gradually more and more Bengali' ' in construction as well as vocabulary and pronunciation. In South-West Bengali- too, Hindi is returned by many of the Hill tribes, who have abandoned their own tongue in favour of a very vulgarised form of the language of the plains. The variety of dialect prevailing >< in the submontane tract and lower valleys of the Hiinatlayas west of Nipal requires a few wdrds of explanation. The term "Pahari," Pahari. or Hill tongue, has been taken as distinctive of the whole- body ; but as the dialects are by no means uniform in detail, the tables show them in three sections, the Western or Panjab, the Oelntral, including Grarhwali and Kumaoni, and the Eastern, which is practically equivalent to , the Nipali dialect of the south-western portion of that State. The title of Mpdli is hot available- for this last, as it has been appropriated by the reginients of N'ipal soldiers collectively known as G-iirkhas, whose language is very largely of the Thibetan type. At the same time it is said that owing to the adoption by "the ruling families of the Khas or Parbatia dialect, which, as has been already^ stated, is based oh Hindi, there is a tendency on the part of the tribes of the higher valleys to abandon their own tongue as they enter more intimately into the military, system of the State. But until Nipal is brought under the census pperatidns we are without clear information on this point. As regards the territorial distribiition of Hindi, it may be observed from the table on page 364 of Vol. I. of the Returns that though this language was returned in every one of the main divisions of India, out of the 85i millions of soi disant Hindi-speakers, 77 millions were enumerated in Hindustan, Bihar, and the Panjab, and seven millions more in the Central Provinces. lie ' In connection with the statistics of Hindi, a few remarks are needed about the (Jrdti. Urdu fbrra of that language. As to Upper India, it may be broadly stated that no line can be drawn between the vernacular and the Persianised speech of the larger towns, where, as before observed, the foreign element has been assimilated to the local construction. South of the Yindhya range, in places where the Musalman element is in a consideraWe minority, and has been recruited in such a way that the converted communities do not retain their original status, there is in use a distinctive dialect, based on the Dehli dialect, largely tempered with vernacular words. This was returned as Musalmani or Deccani, and it has been taken to be the equivalent in the south of the Urdu,-, or Iv&gM, franca of the foreign settlers in the north. It must be distinctly understood, therefore, that the ^^ millions entered under this designation by no means represent the relative prevalence of the diction of the Bagh-o-bahar and the Prem Sdgar. N!ext in order of the Northern group comes Panjabi, which differs from Hindi Panjabi. :?pither in vocabulary and pronunciation than in any other particular. It owes its position as an independent language more to political causes than anything else, and throughout the tract of its adoption shows variety every few miles. /The three main branches are the south-western, called Jatki or Multani,, the Dogri and th(? Pahari. The first is returned by If million p^ersons, all from the plains of the Indus and the Borders of Sindh.* 'The second is the lower hill dialect of Jammii westward, so it is iffe^ft-dsented more in the Kashmer State than in our returns, and is, moreover, lapsing itfto Panjabi,' where it meets the tongue of the plains. The third, which has been libticed' above, shows over 1^ millions of representatives. In the south of the province the langiiage is mixed with B^i-i,- here classed with Marwadi though in the eastern distt-icts the latter element is very weak. 'Curiously enough in the west there is no trace of Sindhi. Owing mainly to the number of recruits furnished to the native arniy by this' province, Panjabi is found repreSe/nted widely over India and Burma, but as a vernacular it is confined to the' lafid'of the Five Rivers. Of ith^. other languages in this group there is little to say. Kashmiri, the most Kasbmeri interesting and inportant, has strayed but a short way out of its native vaUey, and is ^^^^^j^^g, spoken chiefly by ^ the colonies of weavers and carpenters in the Panjdb, and by dialects. "" W3;n(iering shawl-merchants, elsewhere. The falling off since the last census is probably 4ijp/|;ofthe temporary immigration in 1879 of people then allowed for the first time to ^y)M» Ihe tongue of 'the^Jats, Getm (Xanthii), through •whom we derive probabljonr words Egyptian, Gip«y Gitano, as it was a colony , of t.his tribe which was first; transported from Western India to Persia and Asi% Minorj and. spread thence iijto the Lpwer Empire, and through Eiimelia up the Danube. .; ^. . ■''■■ "■■'■'■ S 2 ' 140^ Sindhi. Kacbhi. Marwadi, L?,nguage. Population returning it. Sindhi Kachi .... Marwadi .... Gujarathi* Marathit Goanese, &c. Total, Western Group 2,592,341 439,697 1,147,480 10,619,789 18,892,87.'5 37,738 32,729,920 leave their country under the pressure of famine, and vs^ho have now ref.urned. The small settlement of Kashmeri in the North- West Provinces is a permanent one, consisting apparently of the Hindu or literate class, not of Musalman artisans. There are but few instances in the census record of the languages of the Hindu-Kiish, i In the Panjiib a few families of Grhilghit, Chitral, and even of Kdfiristan and HunKa, or Borishka, origin were found, but the bulk of the population of those tracts was altogether outside the census. There is no need, therefore, to enter into the discussion of the correct nomenclature or classification of this group of littio known but philologically interesting languages. The Western group includes nearly 33f millions. Taking the languages iii their geographical order, Sindhi comes first. Philologically, it is a more backward language than Hindi, having retained far more of the inflections it derives from its Sanskrit parentage. On the other hand, owing to its frontier position, it has borrowed freely from the languages of the Persian Gulf. Indeed, since the country has passed under British rule, a modification of the Arabic character has been adopted in place of the local com- binations of debased Devauiagari letters. *Patnuii, 77,534. fKonkani, 314,485. These last, both here and in parts of the Panjab, only serve to record personal memoranda or accounts, and can hardly be called a means of correspondence, since one of the main objects of the scribe is to keep his production legible by no one but himself. Sindhi is not at all a widely spread language, and most of those speaking- it beyond the province are to be found in the two States, of Kachh and Bahawalpur, contiguous to Sindh, where the proportion of the sexes amongst those returning it shows that the migration is only the ordinary interchange of children in marriage between adjacent villages. Like the Kashmer trader, however, the merchant of Sindh is met with in most of the large towns of India, and, if a census were taken of Eussia, he would be found in the far-off markets of Bokhara and Samarkand, and even in Jj'izhni-r Novgorod, or, rather, he would have been so found at the time of the Indian census, for there is a report that this race has since received notice to quit the territory of the white Tsar. The connection between Kachhi and Sindhi is very close, and were it not tha]b the parent State of the former is politically and geographically more in touch with ^ujardth than with Sindh, it would be hard to say that Kachhi was not still a dialect of the latter. In the present day, however, the vernacular is receiving a strong tinge from its southern neighbour. As Kachhi is the home of a large body of the most enterprising merchants of Western India, it is not surprising to find their language returned from nearly all parts of the country. The case of Marwadi, again, is one in which it is hard to say whether a distinction should be drawn, as with Panjabi, or whether the language thus returned should not be absorbed, like Btmdeli, Maithili, and so on, into the general term Hindi. We have not, however, the advantage of seeing from the census of Marwdr what is the opinion of the people of that State and Bikandr, or rather, .that of their enumerators, on this question. At any rate, Marwadi approaches very near Gujardthi as the latter is spoken on the south border of Kajputana ; so it is taken as forming a link between Gujarathi and Hindi. Like Sindhi and Panj4bi, when written by Hindus, Marwadi has no graphic merits, and several good old tales are current as to the consequences of extending its use to' correspondence, a purpose for which it was never intended. As regards its local prevalence, it is superfluous to state here that Marwadi i-s a widely spread tongue. As we have seen in the second chapter of this work, the thrifty denizen of the sands of Western and Northern Rajputdna has found his way to fdrtim© all over India, from the petty grbcef's shop in a Becoan village to the moBt^extensive banking and broking connection in the commercial capitals of both East and West India. The census returns show us Marwadi as a vernacular only in the tracts immediately adjacent to Rajputdna, but it appears largely in Berar, where the race has established itself in the villages and in connection with the cotton- trade, and also in -Bombay at^d the Central Proviiices, -lor much the same reason,- -vdth additional 141 attraction in the case of the former, of the grand field of speculation in rain, grain and silver afforded by the circumstances ofi the chief town. Relatively to the population, and the distance of the region from their native land, the Mdrwddi is- strong in Assam, but Burma is a country which he has not yet begun to tap. To a peaceful man, whose inclination is to start life in a village remote from competitors, commercial existence in rural Burma is too full of the unexpected to be palatable. ; Gujarathi, in like manner, has become the, commercial language of Western India, GujaratU, But as it is also the vernacular of a considerable area of JBritish territory, as w^U as of Baroda and many of the States connected with the Bombay Presidency, it appears in the returns to be nearly ten times as ' prevalent as Marwddi. In reality, it may be abont thrice as numerously represented, if the whole country be taken into account. Gujardthi is one of the main oflFshoots of the Prakrit of Northern India, and differs little, save in detail and its more complex inflection, from vernacular Hindi. It is almost free from dialect, but as it is the language of commerce, it acquires modifications according to the class using it. For example, the Musalman, trader combines it with Kachi, and the Parsis, who have abandoned their original tongue in favour of it, have engrafted certain peculiarities of their own. One very remarkable offshoot of Gujardthi is found in the Patnlili or Saurdshtri dialect of the silk weavers of the Deccan and Madras. The migrations of this class have not been clearly traced, but • probably it was first brought above the Ghats through one of the many local courts of old time in the Deccan.* The descendants of the original silk weavers are how found exorcising the same trade in Mysore, the Deccan, and in quite the south of the peninsula. The dialect they use is peculiar to themselves, and is not, current amongst them when dealing with other communities, though it has taken the colour of the countries through which the caete has passed, and is at present mainly Telugu, whereby it has lost its northern twang. The reason for this segregation may be found, perhaps, in the fact that a class of this sort, especially when engaged in a lucrative industry, raises its demands for social recognition as it recedes farther from its place of origin. We thus find the Saurashtri weaver of the south employing priests of his own caste, who claim Brahmanical honours and ignore connection with a region where silk- weavers are not in such a high position. This leads them to neglect or depreciate their former tongue. There are, nevertheless, over 77,000 Patniili in the Madras province who still return their language as- of yore. • In addition to the extension given to the Gujarathi language by traders and artisans, there has been a considerable movement up the Tapti valley, in the shape of agricultural colonists, from the plains below, and Sindh, too, has received its share from across the lower desert. Then, too, there is the well-known class of domestic servants, called Siiratis, who were returned in most provinces in India. Marathi is the phief language in the Western group. In structure and vocabulary Marathi. it is ' remarkable for its adherence to the later Sanskrit dialects of North India, and may be called patticularly Brahmanic in all its elements. It has suffered less change than the rest of its companions, and retains much of the complexity of grammatical' form that has been sloughed off by Hindi and Gujarathi. The Brahmans of Western; India have for generations borne a high reputation for scholarship, and as the region under their influence was late in receiving the shock of Musalman invasion, and more or less successful in repelling it, the Hindu character of the language has been well maintained, whilst its fertility, has enabled it to develop without the expedient of continually taking to itself fresh grafts from the Sanskrit. It is what some writers have called a " playful " tongue, abounding in jingles and alliterations, such as "■dndhala-pangala," a blind and lame man; " dagad-gigad," a stone; " lagbag," close by, and "jat-pat," quickly. Foreign words, too, are not respected, for we find " bandhuk-enduk," a gun; " pddri-widri," a missionary; " ard6shi-pard6shi," the neighbours, and so on. The Marathas have named their hill-forts and any villages that have anything striking about their position on the same principle, and their tongue has struck out, it is said, more diminutives and secondary words than any other of the Prakritic derivatives. Setting aside the variety of pronunciation, Marathi is singularly uniform throughout the Deccan. There are, it is true, certain entries * Hofrath Dr. Biibler has directed my attention, since this was first written, to the Gupta InseriptioD, translated at page- ?9, voL iii., of the Corpus Inscriptionum. In this, the colony of silk weavers which immi- grated iojaasapfiJ-CMandesiir; from Central and Southern Gujaratb, are praised for their industry and piety ; ^lieisttei-TaeiDg^ shown by the erectjon of a^temple to the Sun ip tjie time gt" Kumara Grtipta. — J.A.B. 142 HalaSl. from tlie north of that tract giving Ahirani da a distinct language, but this is the resTilt of caste, not linguistic, difference. It may ■ be recollected 'that the later Prakrit grammarians applied the ancient name Apabhransha specially to the' language of the Abhira or Ahir,the cattle-grazing communities of • the aor-th-western plains, and 'there has always been some jealousy between the agricultural classes and those of nomad stock who settle down amongst them. at, a later periodt. A real distinction is drawn, however, on the other hand, between the Marathi of the table-land and that of the Konkan, or the strip of country between the coast and the fbot of' the Grhats. As far south of Goa this distinction ig mot recognised, except amongst a small community of native OhristianSj who entered that fold under the auspices of Portuguese mis- sionaries, and have adopted the language of their instructors from Goa ; for in the latter territory -there is a decidfedly Portuguese element in the vernacular, due probably to the discouragement •of^ native studies and to the destruction of the records in the vernacular when the western Christians first occupied the coast and took in hand the propagation of their religion. The basis of the language, however, is distinctly Marathi. Farther south, the foreign element changes to Kanarese, which is more perceptible as we approach Man^alore, where Konkanl gives way "to piire Kanarese and to TlilTi. But wherever Konfcani is the vernacular it may be held to be a variety of Marathi. The language spoken by' the Christian natives of Groa, who are found 'nearly all over British India^ is usually given as Groanese, but so many have returned it as Portuguese that to avoid misconception the two have been combined, and are placed after Marathi in the Indio,not the European, section of the tables. The return of birthplace shows that oidy 168 perkons were born in Portugal, and whatever may be the language used by the class in question in their correspondence, the proportion of Portiigu'ese ' in the spoken dialect is insignificant. The geographical distribution of Maraithi remains to be noticed. Its covers a wide area, including, as it does, the whole of the North Deccan plateau, the Konkan coast, Berar, the western portion of the Central Provinces, and the greater part of the western districts of the Nizam's dominions. It is met above the Ghats by Kanarese in the south and Telugu in' the south-east. On the coast, as just stated, it merges through Konkani into Karildrese or Tiilil. In the north, it meets Hindi and Gujarathi, without mixing with' either. Settlements' of MarathaS are to be found in Mysore, and there is the small'Tanjore- colony in the south of Madras ; biit, as a rule, the language is not found much in tracts where it is not the vernacular. In Baroda, for instance, which is 'a' Maratha State, there are only about 52,000 Marathi-speakers, mostly confined to the sbuth division of the State, where there is a forest tract in which it is' the vernacular, arid to the troops and retinues of the Gaikw^r and his officials. In other parts of the 'west and centre df India, those who return this language are chiefly in the native regiments, or engaged on railways or as clerks, and in the north as |)riests. Crossing India from the Deccan, only one. Aryan language, is mqt before we entep the lla^tejpu group of that , family. This is.the JpLalabi, w;hich is prevalent only in the south-eastern, portions of the Central Prpyinces and the trac^ts adjaceiil; thereto. It has been grquped with the eastern Aryan tongues from , its position,^although in structure and vocabukry it is inore akin tp Hinfjii In fact, it seems doubtful if it .is ^ot , mainly a, dialect of the latter adopted by the Hill men on coming, intq con- tact with others from the plains, like |;he cases mentioned above, in connection with the (^angetic valley, i On the other hand, it is surrounded by Dravidian Hill languages, and the people by whom it is spoken are not reputed to have made such an advance in pivilisation as would lead to the improvement of their former language, Their whole strength is only 143,000 souls. , i '. Language.' ; ' Population returning it. Halabi - ' - ■ Uriya , i - , ,r . - . Bengali Assamese - - ' - Total, Eastern Group 143,720 , 9,pi0,957 41,343,672 1,435,820' 51,934,169 IJriya. The Bastern group proper begins with Uriya or Utkali, the least advanced of the Prakritic tongues. The. country' of its adoption is diffi-cul't of access, and the hilly tracts that protect it on the west bear an evil name for air and water, the two elements of comfort in a residence that a native of India first values. They "have beeti left accordingly, in tlie possession of a collection of tribes of the raCe either Dravidian or Kolarian. • The language of the Hindu population of Orissa,i therefore, has retained a good deal of its archaic form and vocabulary. It is also peculiarly free from dialectic variation. In addition to geographical isalation^ LTriya has the drawback of a remarkably complicated and awkward character, attributed by several good authorities to the use of strips of the leaf of the Palmyra (borassus flabelUformis) as writing, material, on which the local scribe performs with a sharp. steel stylus, in place of the reed pen common to most of Aryan India. Thus the horizontal line which' keeps together the Devandgari characters and their immediate descendants is here out of the question, as it, [would split thp leaf along its fibre, so the top stroke is omitted. Again, it appears that in writing, the : style is worked on the left: thumb as a fulcrum, a process which imparts a circular form to the results. Mr: Beames, from vyhose work the above J eKplanaition has been taken, adds, "Perhaps the above account may not " seem very, convincing' to European readers, but no one who has ever seen an IJriya " working away with both hands at his style and strip will question the accuracy of " the assertion, and though the fact naay not be of much value, I may add that the " native explanation of the origin of their alphabet agrees with this." It should be mentioned in confirmation of the above that the same explanation has been adopted by Dr. Oald:well, the chief authority on Dravidian language, the southern forms of which are always written on the same material as Uriya. Mr. Beames states also that when he wroije in 1871, the extension of the use of paper was leading to the abandonment of the round top to the Uriya letters ; but judging from the census returns, the fashion is certainly dying hard. The literature of Onsfla is neither extensive nor valuable, except as showing the little change the language has undergone in the last few centuries. As to the geographical prevalence of Uriya, it appears that this tongue is nractioally confined to the tract from which it takes its name and the bordering' districts and States of Madras and the Central Provinces. , Its representatives found at the time of the census elsewhere in India belonged pi-obably to the class of palki- bearers and domestic servants, for which Orissa and; the north-east coast generally _ are famous. Bengali is the largest in point of numbers in this group, and is surpassed in this Bengali. respect by Hindi alone. Though it has no distinct dialects, it resembles Panjabi in having no standard, so that it varies from place to place. more probably than any other of the Aryan tongues. It is true that some years ago an attempt was made to degrade TJriya into a dialect of Bengali, so as to exclude it from schools and public offices ; but fortunately history and philology prevailed over plolitical ambition, and the elder language has held its own against its hybrid sister. For Bengali has no doubt been unfortunate in the circumstances that have attended its development. The latest of' all the;Prakrit ofishoots' to be recognised as a language at all, it dates in that capacity only from the decay of the Dehli empire. Bengal, too, is the province of all others iu which there is the widest gap between the small literary castes and the masses of the people. One of the results is that the vernacular has been split into two sections; first, the tongue of the people at large, which, as remarked above, changes every few miles ; secondly, the literary dialect, known only through the press, and not intelligible to those who do not also know Sanskrit. The latter form is the product of what may be called the revival of learning in Eastern India, consequent upon the settlement of the British on the Hughli. The vernacular was then found rude and meagre, owing to the absence of scholarship ^ and the general neglect of the country during the Moghal rule. Instead of strengthening the existing web from the same material, every effbrt was made in Calcutta, then the only seat of instruction, to embroider upon the feeble old frame a grotesque and elaborate pattern in Sanskrit, and to pilfer from that tongue whatever in the way of vocabularyjand construction the learned considered necessary ^to' satisfy the increasing demands of modern intercourse. He who trusts to the charity of others, says Swift, will always be poor ; so Bengali, as a vernacular, has been stunted in its growth by thisprddess of cramming with'a class of food it is unable to assimilate. The simile used by Mr. Beames is a good one. He Hkens Bengali to an overgrown child tied to its mother's apron-string, and always looking td her for help, when it oiight to be supporting itself. For instance, when the instructions for filling up the census schedule had to be translated into the vernacular for use in this province, the local Superintendent of the operations, a civil servant of much experience of men and cities, obtained' versions, not from the men of liglit and leading in the capital, but from officers administering districts, who knew what the lieges can and will understand. A good working translation was thus obtained, which, when read over to a Calcutta S 4 144 Assamese. Sanskrit. Tami]. scholar, no doubt made him stare and gasp, and mouru over the oppoiltunity -thus lost of giving the widest possible dissemination of culture in style. It is to be feared tfcat not Bengali alone is passing into the hands of oflS.cial scholarship, though this language is probably the least unwilling captive. Bengali, in whatever form it may bejds spread all over the southern valley of Assam, which was till 20 years ; ago a part of the province of Bengal. In Arrakan, too, this language has made some way^ but elsewhere it is confined to a few colonies, at religious centres, and to persons employed in offices of Government or the railway companies, both of which look largely to Bengal for their supply of clerical labour. Assamese, which concludes the list of Arj'^an vernaculars, has had, like iJriya, the experience of resisting the attempt of the ambitious Bengali to reduce it to a patois, and thus open a wider field of employment to the studious youth of the Lower Provinces ; and, like TJriya, too, Assamese has been hitherto successful.* Possibly the political separation of the province from Bengal may have helped on the declaration of independence, for the tongue of the eastern and northern districts of Bengal bears scarcely greater resemblance than Assamese to the euphuistic speech of Calcutta and Nadiya. Assamese is hardly found out of its native valley and the immediate neighbourhood. In bringing to a close this review of the Aryan tongues of India, the fact should be mentioned that 308 persons returned the language habitually spoken in their parents* household as Sanskrit. Tliis survival of 25 centuries is a thousand less than the corresponding tale in the last census returns ; still it is hardly expected that the ladies, at all events, of the households in question spoke in a classical language. The truth seems to be that in the south of India the term Ndgaram is used by the priestly families of the Gujarathi silk weavers to denote their divergence from the ordinary language of their caste, whilst in the rest of the country, the entries may be set down to schoolboys or undergraduates studying the rudiments of the ancient tongue of their faith. In dealing with the Dravidian section of the subject, two groups. have been formed. The first or southern covers nearly the whole field occupied by this class of languages, but the balance, which is here called the northern, comprises tribal forms of speech differing from the rest, except in general structure ; so it has been thought best to keep them apart. The first language to be mentioned is the Tamil, not because it is numerically or geographically the most important, but by reason of its being the most cultivated and the best known of the Dravidian group. It has, indeed, given to the latter its original name.f Two dialecte are by. some attributed to Tamil, but these seem rather caste varieties than those of special tracts ; so both Iruldr and KimUmbar have been included under the general head, whilst Ydrulcala, which has been in: some cases separately tabulated, has been added to the Gipsy dialects, of whieh mention has been made above; Tamil, then, may be taken to be a homogeneous language, covering the whole of Southern India up: to Mysore and the Ghats on the west, and the Ceded Districts, as they are. called, on the north. It is disseminated,, however, as widely as any tongue in India by special classes, such, as labourers, who flock chiefly, to Ceylon and Burma, and donaestic servants, who are found hailing from the Tamil country in every large town and cantonment in India. The Madras servant is usually without re- ligious prejudices or scruples as to food; - - ^ head-gear, or ceremonial, so he can * There ia a tendency, howerer, towards assimilation,. I und^rstajifl,,m the present day.— J.A,B, I Dravid is said by Dr. Caldwell to be a corruption of Tamir. — J.A.B. ' .' . - ' Language, Pcpulation returning it. ' Tamil* 15,229,759 Telugu 19,885,137 Kanarese| 9,751,885 Kodagu 37,218 Malay alam 5,428,250 Tulut 491,728 T6da 736 K6ta 1,201 Sinhalese 187 Mahl 3,167 Total, Southern Dravidian 1 Group - J 50,829,268 G6nd 1,379,580 Kandh (Khond) 320,071 Oraon 368,222 Mal-Pahadi^ 30,838 Khar war - - 7,651 Brahui 28,990 Total, Northern Group 2,135,352 Total, Dravidian 52,964,620 . * Kuiumba, 5,288. f Badaga, 30,656. JKoragu, 1,868. .145 aGComrnddateiliimself to all circumstances, in which, respect he is unlike the northern Indian domestic. ' ' ' Next to Tamil comes Malaydlam,' or the language of the Malabar coast. This is an Malayalam. offshoot of Tamilj and has maintained a close relationship with its parent. Its name is derived from " mala," the locSl term for a hill, 'and an abstract noun signifying possession, like the Persian ddr, so that it means the ", mountain tract." The term Malabar is conoparatively ruodernj' as the last syllable is probably the Arabic for a coaat or roadstead. Malayalam has been closely hemmed in by its neighbours, and, iiideed, the whole tract in which it is spoken has been peculiarly secluded from the foreign influences which have swept over the table-land overhanging it. Those returning this language in Caorg and Mysore are probably settlers just within the frontier, or temporary sojourners who have gone above the (jhdts for the coffee season. In other parts of India there is reason to suspect that, except at the seaports, the entry really means Malabari, a vulgar name for Tamil, originating with the early Portuguese travellers. I come now to Kanarese, the north-western representative of the Dravidian group. Kauaiese. As stated in the introductory chapter, its name seems derived from the Telugu words for blach and country, a title that the denizens of the light soils of the eastern table- land would not improbably apply to the rich western tracts of what is now known as " cotton soil," but other derivations have been suggested. The language itself, however, has a far greater affinity with Tamil than with its neighbour on the east. The only form of speech that can be called a dialect of Kanarese is that of the Badaga tribe on the Nilgiris, called by our early historians the Bnrghers. If this be admitted, it seems not unreasonable to attribute to Kodagu, the dialect of Coorg, a like relationship. Both communities have been shut off by their hills and climate from Brahmanising influence, and seem to have preserved in their speech the older forms of Kanarese, which have been rubbed off that language in the busier life of the plains. The tract held by Kanarese is very compact. It includes Mysore, most of Coorg, and a strip of the coast between the Tulu and the Mardthi. Above the Ghats, it stretches eastwards well into the Nizam's territory, and northwards to the Kistna river. Its neighbour, Tulu, is confined to a small area in or near the district of South TJlii. Kanara in Madras, and it is doubtful if it ever prevailed far beyond its present limits. It has the curious feature of linguistic independence without a character or literature of its own. Dr. Caldwell regards it as a very interesting and highly developed tongue, bearing but a distant relationship to Kanarese and Kodagu, still more distant to Malayalam, which presses on it from the south, and most distant of all to Tamil. The Kanarese character has lately been adopted in printing modern Tulu works, and it is likely that Tulu will ultimately give way to that language. But as it is, it shows an increase of 10^ per cent, over the return of 1881. The last of the chief Dravidian languages is Telugu, which is also the most Telugu. numerously represented at the census. It is said, to be the most euphonious, and, next to Tamil, the most ancient arid advanced of the whole class. There are no dialects returned'under it, for the languages returned as Ydnddi and Chentsu in 1881 are said to be merely tribal corruptions of the standard language, and in no way distinct. On the other hand, the language of certain wandering tribes, such as the Waddar and SaiJcc(di,t}xongh no doubt of Telinga origin, suffers such change in the course of the peregrinations pf these earth-workers and mat-weavers, that in provinces beyond Madi^asit has been grouped with the Gripsy tongues instead of with Telugu, wherever it has been returned under a tribal designation. The Telugu character has been adopted for K^n&rese, as the latter has carried it on to Tiilu. Like the Tamil, Malayalam, and probably the ancient form of Kanarese, it has been modified from the Devanagari of the Southern Asoka inscriptions, curved at top to_ suit the palm-leaf book which is current here as. in Orissa. A very ancient form is still occasionally used in Malabar fpr formal documents, but, for the most part, the resemblance to the original is scarcely traceable. A counter theory has been adopted by some writers, namely that the Asoka character is based on Dravidian forms; but though the question cannqt be considered yet closed, the balance of evidence and probability seems rather on the side of an Aryan ancestry. As a vernacular, Telugu is more widely Spread than Tamil, We find it not only all over the eastern coast of Madras, but throughout the Ceded Districts aboye the Ghats arid in one half of the Nizam*s ' territory. ' It has also gained a considerable I 78388. T 146 Toda and Kote. Sinhalese and MaH. ©ond. Kandh. Ordon and Kharwar, footing in Mysore and in a corner of the Bombay Earnatak, witTi branches in Berar, the Central Provinces, and Orissa. Although only the inhabitants of the North Coromandel coast have any right to the name of Kling, that title is applied on the other side of the Bay of Bengal to labourers from both Kalingapatam and Negapatam. During the rice-shipping season, and now, in fact, throughout the year, Burma is thronged with labourers from this part of the peninsulq,, who have almost monopolized the hard work so distasteful to the Burmese. It was noticed that in some o^.tTae schedules filled up by or for this class, the antiquated term of Grentoo was use^ to denote caste or race. This word was introduced by the Portugiiese, ' to whom all Hindus were Gentiles, even as to the British soldier all natives of India, whether Musalman or not, were collectively Moors. In the present day Gentdo njeans a Telingdna man. There are a few of the minor Dravidian languages belonging to the Southern group which require a word or so of comment. First, the Tdda of the JSTilgiris. Tl^e tribe has received a great deal of attention, ethnological, phrenological, social, and linguistic, mainly because it resides within an easy walk of a favourite hill resort. One member of the tribe has even been distinguished by temporary incorporation into the establishment of the greatest showman of the world-, the late Mr. P. T. Barnum, of N.ew York. Toda is returned by 736 persons, insteai of 673 as in 1881,. and the members of -the tribe are only more numerous by three than those who return the tribal tongue. This last is of similar, origin to the Tamil and Kanarese, and was kept from advance by the secluded life of the tribe in the hills. The K6ta is another of the Nilgiri tribes, lower in position and occupation than the Todas, and speaking a dialect of archaic Kanarese, like that of the Badagas . or Nprtherners mentioned above. Neither of these languages extends beyond the liijiits of the eponymic community^ A few entries of insular languages of this class may as well be noticed here. The Sinhalese is included amongst the Dravidian, sinpe it shows morphological peculiarities which distinguish, it frop^ the Aryan .class in. which, owing perhaps to the cultivation of Pali by the literate classes of the island, it used to be placed. Then, again, the Mahl is returned only by the inhabitants of the little island of Minikoi, situated almost midway between the Maldiv and the Laccadiv archipelago, and belonging "politically, though not linguistically, to the latter. The language in question is allied to the Sinhalese, it is said, though the Arabic character has been introduced,' as in the Laccadiv, to suit thb Musalman proclivities of the people. The Northern group of Dravidian languages comprises one large item, which includes the various dialects spoken by the numerous Gdnd tribes of Central India.' These dialects, which are all unwritten, differ considerably in detail from ea^h bther, but are all Dravidian in their main characteristics. The Gdnd tract, entered in the old maps as " Gdndwana — Unexplored," radiates from the Central Provinces into Bengali Berar, Madras, and the Nizam's dominions, and the tribes' therein are suffering much change, in the course of which their tribal dialects will infallibly assimilate a great deal of the vernacular of the encircling plains. The Kaudh or Khond of the hills of Orissa and the neighbourhood has, like the T6da, attracted more attention than was perhaps his due, owing to his persistent attempts to keep up the habit of human sacrifice as a means of securing a good harvest. The language is localised to his native hills, and the only wanderers found on the census roll are. in Assam, where, if the return be not due to clerical error, the Kandh speakers must be immigrant labourers who have joined a gang of adventurers from Orissa. It is the same with a few of the Kolarian tribes, which have been found even in Burma, though not retaining their tribal language there. Farther north, we enter the region of the Kolarian languages, and the distinction between them and the Dravidian becomes much weakened. The Oraon, however, and the Kharwar may be said to present decidedly Dravidian features. As to the language of the inhabitants pf the hills of Rajmahal, who hold that tract against the Sauthals, there seems to be less information available. PoUqwing such authority as seemed trustworthy, a Dravidian ancestry is attributed to the language of this curiously isolated community. If this be correct, the survival of this relic so far frQm its fellows seems to indicate the fornier, existence of a considerable Dravidian colony between the Kolarian and the mixed races of Eastern Bengal, though the two former" Riay have been identical in blood. 147 , La?t of the Dravidians has been placed the Brahui, though the link attao|aing it Brahui. .^6 the rest is of the thinnest. There are, however, curious features in common which can hardly be ignored. ^Ethnically, there seems to be absolutely no resemblance between the tawny cattle-graziers of the KheMt desert and the squat dark cultivators of , the peninsula or the, mountain tribes of Central India, any more than can be found between t^e ; former and the Mongoloid settlers of the north-east, to whom Signer Pinzi would,^ on linguistic grounds, attach them. The bulk of Brahui speakers dwell in ^^16chistan, a tract wji,ieh-was not enumerated at the census, and the traditions of the jrace as to their migrations have yet to be satisfactorily investigated. The "language of a goo(J number of the Brahui tribes would no doubt -be returned as Baldch, to which their own tongue is rapidly approaching in the frpntier districts, so here again, we , have a|i instance of the obliteration of valuable philological evidence. ; . The Kolarian class, to which our attention Language. i ■ Population returning it. , Santkal* Munda or K61 Kharria Baiga aud jBhunjia - ---, Porwa or'Kur Bhii ■ - - Sawara Gadaba - - - Juang and Maler , . . Total, Kolarian 1 1,709,680 654,607 67,772 48,883 , 185,77:5 148,596 102,039" 29,789 11,965 2,959,006 *.This figure for Santhal has beeu since reduced, on final revi-sion of the Bengal return, to 1,642,154, the balance being distributed apparently between Bengali and Hindi. is now directed, is small, but consi- derably subdivided amongst tribal dialects. The chief is the Santhal, which includes nearly three-fourths of the whole. This language, like so many others, owes its reduction to grammar and writing, to the labours of missionaries, chiefly Danes and Grermans, who have devoted their lives to tlie study of this race. It is widely disseminated over Bengal, as the Santhals are first-rate la- bourers, and show no reluctance to leave their native country for work at a distance. They are found- even in .Assam, both as settlers and as temporary hands on tea estates. Kolarian Class. The Munda, in which returns of K61 and Ho have been included, has apparently Munda and a somewhat wider range, but this is probably due to the ■ partition of their tract allied tribal amongst three provinces. Like the Santhal, the MundA has taken kindly t() foreign 'ii'*'!'"'*^. labour. The language seems . to resemble Santhal in its main features, but its, local varieties are frequent, and each main clan returns a separate designation to both itself and its language, and also, by the Avay, to its religion. The Kharria is on the same pattern. The Baiga, Bhinjwa, or Bhunjia is probably a tribe of earlier date in the country .than the Munda or the Dravidians of these hills, judging from the fact that amongst the other tribes the name is used for a priest or exorcist, even if the filnctibnary is not. of the Baiga tribe. In the Central Provinces, the Baigas are said to spetik Hindi to a lai'g6."extfent, but whether their women do or" not is a more important point, on which; information has not been furnished. The Kur, Kdrwa, or Kiir and Kllrku Ted/ches across the hills to the west of the Kolarian tract', and 'joins the Bhii, Bhii. which carries the class on nearly to' the sea. The former varies in detail throughout it's length, and is diffei-ent in "Berar, for instance, from what it will be found to be in the Chlitia Ndgpiir division of Bengal. As to the Bhii tongue, it has been Adulterated to such an extent that it retains little but corrupt Hindi, Marathi, or Gujarathi, according to lbcality,'in its vocabulary, though its construction, at all feyelits in the eastern portions of the tract it occupies, is still agglutinativa The most southerly forms of Kolarian speech are the S4wara and Gradaba, Sawaru, &c. botli ,qf which are alrhost confined to the HUs bordering on Orissa, in the Madras I^residency. Tihe former may have extended nearly to the sea in earlier days, as it is said,. that the Suaroi.ot Greek geographers are the ancestors of the present Sawara: The last, of this class of languages which has to be noticed is the Juang or Patuii, ^poken by a tribe which Colonel Dalton, in his work on Bengal Ethnology, considers the lowest of all found within tli'e tract he describes. It is but within the last few years that either sex wore clothes at all. The women were till then content with benches of leaves tied round the waist in front and behind, and renewed, as required, when the fair wearer went ' to. fetch in the cattle from the wood which provided her n^illinery.* With the Juaiig is grouped the Maler, which is not returned by more liftaii a few- hundred, persons. *'An attempt was made some years ;ba;ek' to introduce waistcloths, which were distributed gratuitously, but it is Ff poEtpd, I believe, that, as a rule, the innovation did not outlast the material. — ^.J.4.B, , ■ ' ' ' ' T 2 ' 148 Gipsy dialects. Kliasi. Thibeto- 13urman Iauoruii:res. Thibetan dialects. Tlie group entitled Gipsjidialects contains matiy items which are in use only amongst a single caste or tribe. The chief of th&se," the respectable body of carriers and cattle-breeders, called Brinjdrals, is. a community which is spread all over Northern India as far as the iDeccan, where many have permanently settled as landholders, speaking Mardthi only. In the south of the latter tract it gives place to a similar tribe, known as the Lambani, which is the term used, too, in the PanjAb. In both provinces the tribe has an evil repute. There are also the numerous tribes of wandering artisans, such as knife-grinders, grindstone- makers, and matting-pi aiters ; the last of whom alone is held popularly to be of bad character, wherever in India it is found. Perhaps, as formerly with the broom-making trade in England, the facility of the occupation makes people regard it as only a cover for nefarious but more lucrative means of livelihood. Then come the tumblers, rope-dancers, and acrobats, with the castes which unabashedly maintain the name of cut-purse {Ganti- chor), all notorious evil livers, each with its special form of thieves' Latin unintelligible to respectable people. Jugglers and snake-charmers, who are equally nomad in habit, are not invariably regarded malevolently, though open to suspicion. These tribes collectively represent the class from which originate the Eom or Gipsy of Europe and Asia Minor and the Liiri of Persia. Their name of Gipsy is probably a corruption of Zotti, or Jat, as already mentioned, which became " Egyptian," either by misnomer, or because, as some say, Nikomedia, through which they reached Europe, was some- times called " Little Egypt." The other term for them, Zigane, is probably due to their capacity for music, Chang being a stringed instrument in Persia, and Tchingidn or Tsigan, the musician, in Western Turkey. There remains the small group of Khasi dialects (178,637) between the above and the enormous field of the Thibeto-Burman tongues. It has been stated already that this group cannot be aflBliated to any of its neighbours, and it is confined to the range of hills that separates the two valleys of Assam. There are three tribal dialects the Lyngam, Dy^ko, and Synteng. Through the efibrts of the "Welsh missionaries in this tract, the Ronian character is now used in all three, and the census was taken in it. It seems thoroughly established, and as the surrounding languages are different in structure, Khasi will probably withstand their influence. The Thibeto-Burman group, which is the first and largest of the Tonic class with which the census has to deal, has been subdivided, chiefly for convenience in treatment, into geographical sections, in the same way as the Aryan. The first of these comprises not only the Thibetan of the region immediately to the north of the Himalaya, but the language of the Mongoloid tribes of the higher valleys on this side of the range, and those of Sikkim and Bhotdn, none of which regions were included in the census. We are thus dealing with immigrants, except in the com- paratively insignificant case of Kana- wari, or the language of the highfii" valley of the Satlaj, where it passes through the State of Bashahr, between Thibet and the Panjab. The Thibetan was found chiefly in the valleys of the Panjab Himalaya, in the neigh- bourhood of Sikkim and the Bengal Himalaya, and in the Tarai of the north-west, which is the winter resort of the inhabitants of the upper valleys of Kumaon and Garhwal. The Rong or Lepcha is nearly allied to the Thibetan, like the Bhotani, or the tongue of that portion of the Hima- layas known in India a;s Bh6tant, or the end of Bhdt, Thibet. The former language is found only iat Darjiling, and the latter either there or amongst the wintea? traders in North Assam. Most of tlie languages, of Nipal of Language. Population returning it. Thibetan Kanawari Nipali, unspecified „ Mlirmi „ Mangar „ Limbu „ Newari „ Sunuwar Otlier Nipali dialects Lepcha ... Bhotani Total, Himalayan Group B6do (Kachari) Garo Laliing K6ch Mech Tipperah - - - Other Bfido dialects Total, Bddo Group Miri-Abor ... Other Frontier dialects Total, N.^. Frontier Group 20,544 9,265 141,273 20,597 11,281 12,605 5,217 4,236 657 10,125 9,470 245,270 198,70.i 145,425 40,204 8,10^ 90,796 121,864 4,314 609,415 35,703 1,282 36,885 149 Language. ' ' Population returning it. Naga dialects* ... Mikir ^ ■ . . . Kakbyin and Lisliau . . - 102,908 90,236 5,669 Total, Naga-Kakhyin Group 198,813 'J Maoipuri (Kathe) Kuki Zh& (Mskai) Shyin ... 88,911 18,828 41,926 126,915 Total, Ehyin-Ltisliai Group - 276,680 Arrakanese (Magh) Burmese 366,403 5,560,461 Total, Burmese Group 5,926,864 Nikobari 1 Total, Thibeto-Burman Group - 7,293,928 * Seven, but not completely distinguished in the returns. the Tonic class are the mother tongues of people living beyond the British frontier, and mafiy of those mentioned by Mr. Brian' Hodgson, the chief authority on the subject, were not represented at the census at all, either because the tribes bad not penetrated down to India, or, having got there, were confounded by a Hindu enume- rator in the general term Nipali. In Nipal addition to the recruits in our infantry •^*1^<'*«- regiments, Nipdl furnishes a fair num- ber of immigrants of Himalayan race to the adjoining hill country in Bengal and the west of Assam. The G-tirting and Mangar are the main languages returned by the former class, and Murmi, Limbu, and "Newari, by the latter. It is remarkable that we find hardly any trace of the great Nipali family of Kiranti languages in the schedules, and none of those of what Mr. Hodgson calls the " Broken tribes " of the lower portion of the State. The Nipali soldiery as often as not had their language set down as Gurkhali, a title to which probably very few of them would have advanced a claim in their own country. ■ Keeping still to the east, we enter the thorny path of Assam Hill philology, Bddo which requires as caiitious treading as those of the country itself, where stockades, dialects, pitfalls, and bamboo caltrops beset the unwary at every turn. The languages of the western and southern tracts are comparatively easy to classify, as they have been studied for some years. Following the practice of local experts, they have been formed into a separate group, called, from the leading dialect, the Bodo or Eachdri. This last has been fully examined in the works of Messrs. Hodgson and Endle. It is spread widely over the tract in question, and holds nearly the whole of the north-west of Assam. There is no literature or means of keeping up a standard of B6do, and variations are therefore frequent. It is curious that no trace of the allied language, Dhimal, mentioned by Hodgson, should appear in the Assam return, nor in Bengal are there more than a couple of individuals returning it. It is probably confined to the Eastern Tarai, within Nipal territory. So : the two persons just mentioned have been included under the Nipal troup. The Garo language is current in the Western Assam Hills, and is said to be closely allied to the Bddo. The Mech, Rdbha, Hajong, and Hojai, with the Laliing, are all more or less local forms of the same family, and prevail respectively in small tracts along the Bengal and Assam border-land an,d the north bank of the Brdhmaptitra. There are two other languages of this group worth note. The Kdchh, which includes the so-called Pani-Edchh, belongs to an early tribe of settlers along the great river, which has been almost entirely converted to Brahmanism. They number no less than 2,304,000, but the vernacular has been aban- doned for Bengali towards the west, and the small State which represents the former Kdchh domination found no more than a dozen in it to return the ancestral toijgue. The Tipperah language is found almost contiguous to the south-eastern branch of tli0 Bddo, with which it is closely connected. It is returned chiefly from the State bearing this name on the Bengal frontier. The repelling force of the Hill tribes of a more northern descent has proved sufficient here, as farther south, to keep the dialects of the plains strictly within their political limits. Of the curious group of Tonic languages spoken amongst the wild tribes of our Miri-Abor north-eastern frontier, we have but little on the. census record. The Aka, Daphla g^oup. Abor, jyiiri, and Mishmi inhabit the lower valleys of the Himalaya and send but few offshoots into what is known there as the " inner belt " of our influence. There is a settled colony of the Mishmi, however,, in British territory, and some progress has been made with the investigation of the languages, of this and the neighbouring )tribes by frontier officers, to ,^hom we jjave to look at present for our knowledge of the subject. Probablj' better acquaintance will lead tQ tb§ discpv^jy of more linjis between T 3 150 Naga- Kakhyin group. Kuki-Llisliai group. the various farms of speech amongst them; for in this corner of the empire tribal hostility is the leading feature of intercourse, and the mere fact of. separation is reason enough for considering a neigkbpuJ" 9- natural enemy, and in disclaiming, of course), any former connection with him. Between the above group and. the next in order lies a small wedge of the Tai or Shan family, which will be deailt with in connection with the languages of Burma later on. The Naga group is not confined to the tribes which bear that name, but, as stated in the introductory section of this chapter, includes the inhabitants of the Mikir hills to the north-west, and the tract stretching east, probably along the whole of the northern frontier of Upper Burma, which is in the hands of the people collectively known as Kakhyin. There are seven tribal languages of the Nagas returned at the census, but unfortunately there is also a large body of people who are shown as speaking N^ga only, without the distinction that could easily have been obtained had the enumerators been warned to look out for it. We have, for instance, a, fair approximation to the number of the Angami and A-o Nagas, but the Kacha or Empe-d, the Kezhama, the Lho-ta, the Sema, and the Eengama are but poorly represented. The language of* the main . Naga tHbes on the western slope of the Indo-Burman watershed has been reduced to more or less grammatical form by Mr. Davies, the ofl&cer in charge of the Hill tracts, who is at present the best authority on the subject i; and when similai; work has been accomplished for the language of the Burmese tribes of the same race, the two results will be a very valuable contribution to philology. The curious oflF-slLOOt of the Kakhyin race in the north-east corner of Assam has been alre,aidy mentioned under the name of Sing'ph6,.or, more correctly, Qhing-pau. It ha-d its centre ^t Mogaung on a tributary of the Irawadi, near which, the fighting took placfe last year, and established itself in Assam near the end of the last century, and made slaves of a good many of the Assaiaese of the neighbouring tribes. A mixed race, the oflspring of this connection, is in existence, but returns thjO language of the country, not of the foreigner. The whole community is very small, just over 2i000,in all, of whom two-thirds are Singpho, and the regt, Duania, of half-breeds. The next group of languages on which comment is required comprises the languages of the tribes of the range of hills separating India from Burma. In their northern extension these tribes are collectively known as Kiiki. The term Liishai, which is applied farther south, is not recognised by the people themselves, who use the name Zho. Shendti is also a synonymous title for the Liishai tribes. The tribes of the country between Bengal and Burma are known collectively as Kbyin in the east, and by a variety of local names in Bengal, The whole lot was left very much to itself in former years^ as the inhabitants of the plains hold such races in considerable respect, and, trading on this feeling, the mountaineers have manifested their ^periority over the peaceful communities . they overhang in ways that the British had to stop with some vigour. It is hardly necessary to point oui that with so many tribes close together, each under hereditary obligations to lay by a store of the skulls of its neighbour, tTie diversity of language is as great as in the tract across t^ie Brahmapljtra. Information on these dialects tas not yet been obtaineidj'as the hi^ls hg,v6 but recently been permanently occupied. The only civilised community is that of the valley of Manipui-, wjiich acquired so painful a notoriety from the disturbances of March 1891. Tie language of this tract is called Manipuri in Bengal and Assam, and Kathe or Ponnu in Burma. The returns available relate only to the colonists or visitors in the adjacent portions of Assam, and the entourage of. the late chief in Bengal and Mathura. In Mandalay there is still a considerable colony, descended from prisoners or slaves brought over during the wars of the late dynasty. In Lower Burma the number is decreasing, owing either to cessation of immigration or: to, the adoption of the Burmese language by the settlers, since there is no single colony in that part of the province like that in Mandalay, where the similarity of fprtui^e and intimacy of intercourse has helped to keep up the original language. The cultivated dialect of Manipur has been studied and deebribed by Colonel Lewin and Mr, Brojonath Shaha, according to whom it shows a great advance on the rtest of the tribal dialects. It is said to have had a special character of its own, which, like that of Malay alam, is now only uSed occasionally in formal deeds. For ordinary pui*po8es it has been superseded by Bengali, introduced by thfe ofl&cials and settlers from that province. Manijiur, like Nipdl, has ' had its Brahman invasion, with its concomitant traditions of the'o'rthodox descent 'o¥ the ruling family from the demigods of Bindraban on the Jamna, and there is a good deal of mixed blood in the central valley, which accounts for the relatively high -culture of 151 that tract. The hills are still held by Kiiki and other trihes of the same atook, and it is doubtful, therefore, if the language of Imphal.is current far beyond the immediate influehce of the court and its Hindu ojaoers. Between the Zh6 country and Arrakan the tribal languages are badly mixed at Khyin the census. We have, however, grammars and vocabularies of the dialects current on '^'*1®'^**- the Burma side of the Ydma hills, i amongst the Southern Khyins, but the literary Burman, in his contempt for the men of the mountain, avoids discrimination of the rest of these tribes, and to him all Khyins are one, or nearly so. We find, however, a fe'?^ tribal languages set down, ■•such as Mr6 (15,891), Khwe-mi (14,126), and Daignet (856), from Northern Arrakan, but some of the races returning these are doubtless more numerons than the above figures would imply. Possibly, therefore, as the. tribes expand beyond the shelter of the hills, they are absorbed linguistically into the gulf of Burmese. We then come to Burma itself, and' here it has been found advisable to show the Arrakanese. main language iinder the two heads of Burmese and -Arrakanese. Historicallv and philolo'gically they are the sanae, but geographically it is possible to distinguish them. Arrakanese is the older form of the language, and has preserved features, especially of pronunciation, which the soft and indolent inhabitant of the rest of Burma has been willing to let drop. But there are stiU dialects outside Arrakan which belong to it' rather than to the inland tracts. Tavoyer, idr example, the special dialect of Tenasserim, and Ghdungthd, a vanishing reHc, may be cited in proof of this. On the Bengal side of the bay, north and east of Chittagong, the language is returned as Magh, a name uakiiown to Arrakan, and, like the Goanese of the western coast, (iailing to mind the modern kitchen rather than the ancient kingdom from which it is derived, for the Magh and the Groanese furnish European India with the best of its chefs. ' ' < . , , .' Ih Burmese We have the language most numerously represented at the census Burmese, of all the 'Tonic clasfe. , We have jiad occasion.to njention already the tradition that this &c. tongue came from ihe west, but, in 'fact, the only impression received from India via Ceylon is. the alphabet in a very modified form, and a comparatively small' vocabiilary relating to the ' Magadhi-horri religion. ' On the other hand, the appearance of the people betokens a far more northeirn Crigin,.. Within historic times Burmese has proved'itself a great absorbent of local dialects, either owing to political influences from'Pegn and Mandailay or to the prevailing ecclesiastical system, which includes monastic "training and education, Tefem, a tongue held ' to be distinct, from the Burmese, and which was represented by 436 persons in 1881, has vanished froin the returns,, though there are nearly 2,300 of the tribe left. Kddu,Ddm, arid Yau, too, are well on the way to the same. fate, and show signs of decrepitude in the last 10 years, though the tribes are flourishing. But however .stroDgi. the Burmese may be in its own country, it- spxeads hut little into India. Withvthe- exception of the Maghs of Chittagong,-most of the Burmese speakers who are to be found across the sea have strayed thither against their will, and include the late King and his establishment, and the "true paliriots " sca.ttered amongst the Indian jails — . , , - ■ " For^e it understood, They le^t their country for their country's good." The M^n or Talaing languages ;-ma;y- be mentioned here, though little more need M6n group. ' be said about them than is to be found above in connection with the general subject of classification. The la'nguage is found in Anam and Cambodia, though in the latter case the influence of Chinese is very marked. It is worth noting, too, that the same tradition as to the advent of a prince from across the Bay of Bengal is current in aU three communities.; The only link between Pegu and the Mdn colonies of the far east is found in Palaung, a language spoken by a small and isolated tribe in the north-east of Upper Burma. In the time of the late dynasty the M6n language was almost obliterated in Lower Burma, owing to. the d&a^bttk&ement of it by the Burmese rulers and their officials, and it was expected to become extinct in a generation or so. Now, however, the census . returns seem to T 4 , Lauguage. Population returning it. M6n - - - .Ealaing , - - - - Total, M6n Class 226,495 2,847 229,342 152 Language. Population returning it. Shan Ldo {Htai) Ait&n Khamti . . . - Phakial Total, Taic Class 174,871 4 2 2,945 625 178,447 indicate that this language is either reviving, or that persons who were j^eluctaijt to return it oh former occasions have taken heart, for there were only 154,553 returning it in 1881 : but it is imprudent to •prophecy that it will ever resume its position as one of the leading vernaculars in Amherst and Pegu. As yet, it is returned by little over one-half of the Mon-Talaing community in Lower Burma. ghanorTai Between the two branches of the M6n there has been thrust a long belt of the group. Shan language, called in the centre and south Ldo, or, by Europeans, Siamese. It has already been pointed out that the Ah 6m or Shan settlers of Upper Assam failed to preserve their language in their new circumstances, so we find no entries of that tongue amongst them ; and the Ait6ns, a more modern tribe, are re- turned as mostly speaking Shan only, which is very likely due to mistake on the part of Hindu enumerators. The two other Shan colonies in Assam return a few entries of Khamti and Phakial respectively, but the tongue does not extend beyond the limits occupied by those tribes. Khamti has been investigated by a frontier officer, and a grammar written of it ; but there is a gap between this tract and the Lao or Htai of the Siamese kingdom, in which the intervening linguistic changes have still to be investigated. This cannot be done until the ois-Salwin Shan States have been brought into closer touch with the new administration of Upper Burma. By next census, therefore, more information will have been obtained, even if the States be not completely enumerated. Malay group. The Malayan class of languages is represented chiefly in the south of Burma, where there are a few settlers from the peninsula. The Sea- Gipsies or Sal6n of the Mergui Archipelago, too, speak a tongue of this class, and owing to more careful enumeration, show nearly double the number that appear in the returns of 10 years back. The rest of the Malays are seafaring people, enumerated at the ports of India. As this class of languages does not really prevail within the sphere of the census operations, no more need be said about it. Japanese. Japanese can be disposed of in the same way. A few residents of the chief towns sire all that return this language, and as the women returning it are more numerous than the men, its introduction is probably due to the fame of the Musme of the baths and tea-houses of the Far East, which has been productive of emulation. Chinese. The Sinitic class is represented by two languages, Chinese and Karen. The former is scattered nearly all over India, since merchants, carpenters, and carie- workers of this race are to be found in most lairge towns and cantonments. In Eangoon and Upper Burma there are regular communities from the Flowery Land, though in the latter the number of immigrants seems to have decreased since the country became British. In nearly every district, however, there are a few Chinese. They belong mainly to two classes. First, the merchant who hails from Canton or the east, and who comes round by sea; secondly, the people from Yiin-nan, who come across by land to trade in petty wares and forest produce, and whose language is different from that of the former. Karen. Kdren has been classed with Chinese on the authority of Dr. Gushing and other missionary scholars. The Karen tract has been for some years the main field of missionary labour in Burma, and the results are to be found in most districts of the lower division of that province. The three chief dialects of this language differ, eoasiderably from each other, owing probably to much the same reason as is found operative in the case of the Lushai group, namely, the dispersion of the tribe over Language. Population returning it. Malay Sal&n Javanese Total, Malay Class 2,437 1,628 19 4,084 Language. Population returning it. Chinese . - . - Karen . . - - Total, Sinitic Class 38,504 674,846 713,350 153 Language. Population returning it. Persian Armenian - , - Pakhtii (Pashtii) - - - Bal6ch Total, Eranic- Aryan Class - 28,189 . 833 1,080,931 219,475 1,329,428 wide tracts of difl&cult country, after^ tbe first arri^yal from tlie north. The Pwo dialect was returned by .449,450 persons, though only about 27,000 are giyen as of th fit tribe, ^he Sgau ^hows 225,193. The Bghai or Kdrdn-ni was returned un(3er these titles by no more tha,n , 16, so is evidently included in the fi,r8t-named dialect, Taungthu, M6pgha^ and so, on, are said to be only local names for one or other of the above three dialects. The Kdrdn does not extend much beyond the tract pccupied by the tribes themselves, and some of the outskirts are said to be getting rapidly Burmariised in speech. We can here ta-ke leave of the languages of India, and discuss the return of other Eranic- Asiatic iiongues. First comes the sister Aryan Ian- brarich of the Aryan class, which may be Stages- called the Eranic, because, wh-atever its origin, the few representatives we have to deal with are well under the influence of Persian. As regards this last, in Persian spite of the connection of India with Iran, and the settlement of natives of the latter in the north and west of the country, the language has obtained, as a parent tongue, little footing within the empire. It is still the, language of polite society and of belles lettres amongst the educated classes of Musalmans, but those to whom it is the only or the most familiar language are but few. In Bengal and Rangoon there are remnants of the old ruling families of Dehli and Lucknow ; in the Panj^b, traders and immigrants are found, and the refugees from Afghanistan ; and in Bombay, horse-dealers and emigrants from Persia, who have settled down in the chief towns. Beyond these centres there is ha,rdly any real Persian spoken, and a good deal of what is returned as such is but the better sort of Urdu. Armenian is the parent tongue of a small colony of that race in Bengal, Lower Armenian. Burma, and a few other parts of India. The persons returning it are permanently domiciled in India, and probably use English more and more each generation. Pakhtu or Pashtu, the language of the Pathan (Pakhtwan), belongs, as has been PakMii. said, to the Indie- Aryan class in its structure, but to the Eranic in much of its vocabulary. It has been grouped under the labter, as it is geographically beyond India, and Persia,n of a sort is undoubtedly the vernacular of the north and wes+ of Afghanistan. Most of the Pashtu- speakers are located in the Panjab, either as settlers or amongst the retainers of the refugees. There is another class which travels over nearly all India peddling rugs andfruit, and sometinies selling horses brought down from their native land. They leave the latter in the autumn, perambulate the' gehial plains of India during the winter, and return to Grhazni, or wherever it may be, in the late spring, in time to resume theii* home occupation of agriculture and fruit-growing. Year after year they do this, tod many tribes of the Sulaiman and Kharoti map out India into reo-ular divisions, one of which is assigned to be exploited by each company. There are a few who do not return at all, but remain all the year round in India, varying the monotony of commerce, according to the police, with exploits less legitimate. Baloch,' again, is a language of Persian affinity, used as a vernacular only in the Bal6ch. frontier districts of the lower Panjdb and in Sindh. The prevalence of this tongue, according to the returns, in some of the native States, is due to the number of mercenaries of Bal6ch race entertained by the chiefs as personal retainers and treasury guards. They are usually Makranis, speaking that dialect of Bal6ch. They have a strong strain of African blood, and are not of the pure type of the north. Next to Bran, we naturally come to its enemy, linguistically known as Turan, Turanic but the latter is very poorly represented languages, in India linguistically. The majority of Tiirki speakers come from Turkestan, and there are a few Osmanli settled in Bombay and Haiderabad. It is not worth while for so small a number to distinguish the various entries of Uzbegi, Yarkandi, Ghagatai, &c., from the culti- vated dialect of Stamboul. The Firin fentries are from the vessfels in port on Language. Tiirki Finn Magyar Total, Turanic Population returning it. 607 10 42 659 I 78388. IT 154 Semitic languages. Arabic. Hebrew. Negro dialed s. European languages. Language. Population returiiing it. Arabic Hebrew Sjriac Total, Semitic 53,351 2,171 12 55,534 the census night at Calcutta and Rangoon. Magyar is returned by a few merchants and travellers, and by artisans on the railways, where nearly all the tongues of Europe are to be found. The term Turanic has been reserved for this class of languages, instead of being extended, as is sometimes done, to the whole of the Tonic and Agglutinative languages of Burma and the Himalaya, for which the title selected above raises, it is to be hoped, fewer questions. The Semitic is the last of the Asiatic classes with which the census is concerned. It is made up mainly of Arabic, the vernacular of Aden, which Settlement furnishes nearly half the return. Most of the rest of the entries are returned from Haidrabad, where a portion of the Nizam's force and the body-guards of his chief nobles are recruited from Arabic-speaking tribes of Africa and the coast of the Persian Gulf. Hebrew was returned by a number of Jews, much on the name principle that Sanskrit was put down by Brahmans and Arabic by Musalman teachers of the Kurdn. It is the distinctive language of their faith, though it can hardly be called a mother tongue, so it is considered dignifying, if the head of the house is at all conversant with it, to have the fact recorded. The entries were chiefly from Jewish colonies where Arabic, Marathi, or Malayalam would be the correct return. Turkish Arabia supplies a good number of Israelitish settlers. In Cochin there are two permanent colonies, one of which is certainly of foreign origin, whilst in the coast district round Bombay the class of rural Jews habitually use Marathi in their households. In Aden, too, the colony is permanently settled, and speaks a dialect of Arabic. No definite information has been recorded about the few entries of Syriac which are found here and there. Possibly they are traceable to Nestorian priests, of whom there may be some in Cochin and other parts of the country. From Africa we get a number of dialects, found mostly in Aden, where there is a constant stream of migration to and from the opposite coast. The Somali and Swahili tribes furnish most of the entries. The negroes who are employed on board the steamers trading with India are often called Sidi or Habshi, which, strictly speaking, is Abyssinian. They are from farther south, however, and usually take ship at Zanzibar, or come from the Persian Gulf ports, such as Mascat. It would be incorrect, therefore, to class their languages as Abyssinian, which would bring them into the Semitic class. It may be remembered that in the early days of operations against the slave trade on the east coast of Africa, vessels full of boys and girls were captured, and the children made over to certain missionary orphanages and other institutions in Bombay. They were trained to a mechanical trade or to domestic service, and thus remained in the country of their adoption. Another class found in the households of chiefs and rich nobles is still recruited in less reputable ways, mainly on account of the high character for fidelity to their employer borne by men of this race. We have now to consider the return of the languages of Europe and the west. Of these there is a great variety, though none but English is strongly represented. This last, which, for census purposes, is held to include Scotch, though not Gaelic, is returned by 100,000 or so born in the United Kingdom, by scattered denizens of Australia, the United States, and. Canada, &c., as well as by the increasing class of Europeans of British descent born and domiciled in India, and the Eurasians. Of the other languages returned, German, French, Italian, and Greek may be taken as those of commercial sojourners, and "Welsh, Gaelic, and Irish are recruited from the regiments that happen to be on service in India. In the same way, the Scandi- navian languages depend on the number of vessels in port on the census night fqr their representatives, Language. Population returning it. Teutonic 'English - German Dutch 238,499 2,215 119 Flemish •22 Danish 94 Scandinavian -■ Swedish 187 Norwegian '■Welsh 152 245 Keltic - Gaelic Irish 264 299 Keltic (unspecified) r Greek 2 380 Latin 1 Italian i - 690 McditeiTanean -i Maltese 32 Koumanian 22 Spanish French r'Russian -• 159 2,171 95 Polish 46 Slavonic -< Czech - Bulgarian ^Slavonic (unspecified) Basque , - 1 49 1 1 Total, European 245,746 155 Compiaring the return of these languages with that of the preceding census, it will be seen that the Europeans have increased by some 37,000, out of which number more than 35,000 are amongst the English speakers. Changes in the garrison, as well as the natural increase amongst the domiciled population and Eurasians, sufficiently account for the latter. Of the foreigners, the three commercial classes, German, French, and Greek, have all largely increased, but the Italian-speaking population has fallen off, owing partly to the disappearance of a colony of that nationality from Rangoon, where it was established during the existence of the kingdom of Upper Burma, The Scandinavians were much fewer, and the Celts many more, than in 1881, owing to the temporary causes just mentioned. Looking up some of the details, it appears that the priest who returns Latin as being the tongue of his mother, the Church, has safely weathered the intercensal decade ; so, too, have the two Kelts and and the single Slavonian, who still refrain from further particularising their mother tongue ; but the Lapp has given place to a Basque, and representatives of Bohemia and Iceland have appeared, besides various Bulgarians. No general comparison between the returns for the two enumerations has been Comparison attempted, since, partly owing to the wider scope of the present operations, partly of Tribal perhaps, to the better understanding of the rules by the enumerators, it is only in the '*^?"^°^b i purely localised or tribal dialects that such comparison could be practicable. But population, there is a certain interest, connected with what has been said regarding the prevalence of tribal dialects, in comparing the return of languages in this table with that of the tribe in Table XVII. The following statement, accordingly, gives specimens of the results of this comparison, the population where language was not returned being deducted in all cases : — Numerica I Strength. Difference of B. from A. Numerioa Strength. Difl B. Tribe, &c. A.— By Tribe. B.— By Tribal Language. Tribe, &c. A.— By Tribe. B.— By Tribal Language. ereuce of from A • 'Miri-Abor 36,738 35,703 _ 1,035 -Santhil 1,494,045 1,642,154 + 148,109 Kh4si 172,150 178,637 + 6,487 Bhi'l 933,512 148,596 — 784,916 Kaohiri 243,378 198,705 _ 44,673 Baiga, BhiDJwa, &c. 1,030,332 48,883 — 981,449 Kdchh 2,364,365 8,107 2,356,258 ¥ Miindi, H6, K(51,"l Kur, and Korwa J 1,109,157 840,282 - 268,875 a Garo 150,227 145,525 - 4,702 ^ Eharria 68,425 67,772 - 658 V M&h 96,873 90,796 - 6,077 Kharwar - - - 112,298 7,651 - 104,647 mga. 101,568 102,908 - 1,340 Manipuri 84,540 88,911 + 4,371 'Oraon 523,258 368,222 - 155,036 Kuki 25,940 18,828 - 7,112 Gond 2,897,591 1,379,580 - 1,518,011 Liishai 43,840 41,926 - 1,914 1 Sawara 438,317 103,039 - 336,278 "Murmi 21,889 20,597 - 1,292 Gadaba - 34,127 29,789 - 4,338 Mangar 19,383 11,281 - 8,102 1 Eandh 627,388 320,071 - 307,317 t. Limbu 15,079 12,605 - 2,474 Mal-Pahadia 18,506 30,838 + 12,332 SunuwAr 5,210 4,236 - 974 Juang 9,179 11,171 + 1,192 ^Newar 4,979 5,217 X 238 'Burmese 5,408,984 5,560,461 -1- 151,477 'Irala 58i503 1,614 - 56,889 Arrakanese 452,164 366,403 - 85,761 i. T(5da 739 736 _- 3 |. Karen 540,876 674,846 + 133,970 g Kdta 1-,201 1,201 — w Mo'n 467,885 226,495 - 241,390 . - Chinese 41,832 38,504 — 3,328 The general drift of the statement is to show how the tribal tongues are falling before, the general dialects of the plains. This is marked out very plainly in the figures for BMl, Gdnd, Baiga, with its companions, and Kharwar, in the case of Hindi, and equally well in that of , TJriya, amongst the speakers of Kandh and Sdwara. A U 2 ... 156 Table showing, proportionally, the Languages prevalent in each PfbOviNCE, &o. . Province and Language. Per-centage on Total. Province and Language. Per-oentage on Total. 1. Ajmer ;— 12. Upper Burma:— 1. Hindi 56-58 1. Burmese , -. 92-91 2. Marwadi 42-20 2. Shan 2-68 Others I -zz Others - ' - ' - 4-41 2. Assam:— 13. Lower Burma:— 1. 'Bengali 2. Assamese 3. Hindi 4. Kachari - S.Khasi 6. Garo Others 50-06 25-82 4-20 3-59 3-25 ' 2-19 10-89 1. Burmese 2. Karen - ' ■- 3. Arakanese,,- ,j' ,- 4. Mdn -„, , 5. Bengali -■ 'i t-^^ - '6. Khyin • - , ^ , , 7. Shan ' ' Others 60-49 14-28 7-41 4-85 3-50 2-15 2-02 5-3o 8- Bengal:— .14. Coorg:— 1. Bengali - 52-8.9 a. Kiinarese 43-99 2. Hindi 36-65 2. Kodaau 20-40 3. Uriya 'i> 6-46 3. Tamil 9-36 4. Santhal 1-99 4. Malayalam 6-47 Others 2-01 5. Tdld 6-94 6. Urdu 3-99 4. Bengal States:— 7. Telugu .... Others 2-14 6-71 1. TJriya 45-21 2. Bengali 19-72 15. Mysore:-- 3 Hindi -' 15-29 ; ' ■ ^ ' t 4. Santhal 5. Tippera 6. MrtQda^&c- Others 8-25 3-34 3;19 5-00 1. Kanarese 2. Telugu 3. Urdu 4. Tamtl- -- =^^ - 73-94 15-19 4-73 ^'22- Others a-gz 5, Berdr :— 16- Madras:— 1. MarSthi - 79-46 ! ■■'■■ 1. Tamil -^ ' - '. ^ '' 39-51 2. Hindi 9-67 2. Telugu > - 'r ■-!'.' 38-32 3. Gdnd 3-24 3. Malayalam 7-54 L_Ojhers _ _ - 7-63_ _ 4. Kanarese - - ■- 4-06 5. Urjya - - -' 3-63 s. Bombay ;— 6. Urdd ■ .7. Tu ii, ., . - - - 2-29 1-34 ? . Marithi ' ' - " 53-82 : Others i- 3-3i 2. Gujarathi 20-78 '■'■'.,'-- 1 ; , ' 3. Kanarese : - 15-59 17. Madras States:— ' 4. urad ; Others - 5-62 4.-19 1. Malayalam 2. Tainil 73-54 22-86 3. Telugu 1-54 7. Bomliay States:— Others z-06 1. Gujarathi 2. Marathi 60-68 22-19 18. Haidrab^d — 3. Kanarese - 7-25 1. Telugu , - . ! 2. Mair^thi 43-61 4, Kachhi 4-73 30-28 5. Urdu 3-20 3. Kabarese - 12-58 Others 1-95 4. Urdu Others 10-38 3-iS 8. SincLh:— 19. North-West Provinces :— 1. Sindhi 2. Bal(5ch;_ 83-00 6-20 1. Hiiidi , . , 97-06 3. M&rwadi - Others 4-74 6-06 2. Kumaom - ' 3. Gai-hwali Others 1-46 1-19 0-Z9 9. Baroda:— 20. N.W. Province States :— 1. Gujarathi 93-00 1. Hindi 69 ■ 58 2. Urdii -^ S-39 2. Garhwdli 30-32 3. Marathi 2-11 Othe7-s , - . ! 0*10 Others i-So 21. Panjab-.— 10.? Central Provinces:— i 1. Panj&bi ' 63-13 17-63 1. Hindi 60-26 2. Hindi i - 2, Marathi 19-61 3. Jatki ; 8-43 S. Gdnd 9-32 4. Pakhtu - 5-06 4. Uriya - • 6-35 5. Western iPahAri 3-54 5. Urdli 1-48 6. Bagri (M'drwddi') 1-50 O^/ters • i •■ - 3-04 Others ; - _ . 0-71 1 1 . Central Province States :— "' 22. Panjib states:— - ] . TJriya 42-41 1. Panjabi - , , . ,r , - 60-40 2. Hindi 36-06 2. Western Pahdri s . . 18:39 3. Gdnd 8-47 3. Hindi - ■ ■ . '■ '." "'" 11-21 4. Halabi 6-48 4. Jatki - , ' . : . .^•97 5. Kandh Others 3-06 3-52 5, Mar-v)radi (_Bagri) ' Others '- '^' :. ' - ^4 5-61 1-12 157 few items seem to be capable of special explanations. For instance, the Kh^si speakers include those returned as Christians, not under the tribal name. The Manipuris, too, use the title of Kshatria, or the warrior caste of Brahmanic society, to some extent. Assamese, moreover, seems gaining ground amongst the Kachari tribes and their fellows, as Burmese is further south. The discrepancy between the two returns of Karens is not explicable on the surface, as the Christians who have returned that as their race are included for the put-poses of this table under the tribal head. The Chinese have the peculiar custom of attributing their own nationality to the male children of mixed parentage, but of returning the females as Burmese, a fact which seems to account for the difference noticeable in the table. The Irula tongue is no more than a rude dialect of Tatnil, so there is no reason to expect its return under a distinct head. In the case of the Nipal tribes the mother-tongue was mostly entered simply as Nipali or Pahari. The figures for Mal-Pahadi are clearly confused with those of other tribes. There is no more detailed explanation available as to the rest of the discrepancies in the statement. ' The third part of Table X. in the Imperial series shows the linguistic distribution of the population of each province and State. A proportional sutamary of it is shown opposite. It seems to require no special 6omment. B. — The Distribution of the Population by Religion. " One religion after another may fade away, but the religions sense which created them all can never become dead to humanity." — Richter. The various forms of religious profession current in India bear, as a rule, very distant relationship to the ■ ethnid distribution of the population, as each includes recruits from nearly every section of the community. But all bear traces of the general influence mentioned in the opening section of this chapter. That influence, indeed, finds perhaps its strongest manifestation in the hold still universally retained over the beliefs of the masses by concrete notions regarding the relationship between man and the supernatural, such as are associated with the very earliest stages of organised Animism. The subject is, from its very character, too wide to be discussed- here, but it is impossible to deal adequately with the statistics given in Table Yl. of the Imperial series without considerable explanation of the scope of the nomenclature adopted, and this entails something in the way of a disquisition on the development of the various component elements of the general titles which have been thus brought into the service of the census. To begin with a term just used above, it is necessary to show what is here meant Animistic by Animism. As the word is extensively, used by Dr. Tiele in a sense somewhat wider religion, than that which other writers on the subject have given to it, his own interpretation, which is that adopted for the purposes of the present work, may as well be quoted : — "Animism is the belief in the existelice of souls or spirits, of which only the powerful — those on which man feels himSelf dependent, and before which he stands in a>ve— acquire the rank of divine beings, and become objects of worship. These spirits are' conceived as moving freely through earth and aii-, and, ' either of their own accord, or because conjured by some spell, and thus under compulsion, appearing to men (Spiritism). But they may also take up their abode, either pfermahently or teinporarily, in some object, whether lifeless or living it matters not : and this object, as endowed with higher poweu, is then worshipped or employed to protect individuals or communities {Fetishism). Spiritism, essentially the same as what is now called Spiritualism, must be qarefully distinguished from Fetishism, but can only rarely be separated from it." Such is the general conception of Animism, but, as we have no occasion to treat of it in its more primitive forni, it has been used above vith the addition of the word organised. By this is meant that in the stage of development in which we find it in In,dia, it has passed from its archaic and indeterminate shape into a collection of polydaimonistic tribal religions,, in which spells,, magic, and exorcism are all prominent. It is relevant to what has to be said a little later, to point out that in this Stage the malevolent spirits are considered the more important, and little notice is taken of the good. The main object, in the first place, is to get power over the spirits by magic, and, in a higher stage of belief ^ to propitiate them by gifts or homage. These notions are to be found in every stage of evolution in different parts of India, but the term Animistic has been restricted^ in the census tables, to a certain class of the community. As in the case of several other forms of creed in that country, it is necessary to define the scope of the title, negatively, or by explaining what is not included under it, rather than what is. In anticipation, therefore, qf ^explana;tionof,,the term Hindu, it may TT 3 158 r be stated that it was the intention of the framers of the rules for enumeration : that under the head of Animistic should come all niembers of . the forest tribes who were not locally acknowledged to be Hindu, Musalmdn, Christian, or Buddhist, by religion. No general title for the religion of such communities was prescribed, > but the r enumerator was instructed to enter in the column reserved for this information the t the name of the tribe, and the compilation under the above-mentioned head was made at the Central office of the Provincial census. The distinction between the tribal form of faith and that of the lower grades of the Brahmanic community is very elastic, as will be shown below, so that the application of the rule on the borders of the hill tracts was no doubt arbitrary. It is well known, too, that the process of Brahmanising, has been long and continuously in progress amongst these tribes, where, indeed, the? latter religion is said to find the only fields left for its propaganda, so that very probably the more advanced branches' of one of these tribeSj who presumably dwell in closer proximity to the plains, are in fact > affiliated to the Brahmanic faith, which , makes but small demands upon their capacity for change, whilst their fellows in the: remote valleys and forests continue in their former unregenerate condition. Ini Lhis way we can account for the fact that out of 15,922,000 members of the forest tribes, only 9,280,000 are returned under their tribal form of religion. But the distin,ction , is really of little moment, because, as was mentioned in the beginning of this section, evei'y stratum of Indian society is more or less saturated with Animistic conceptions but little raised above those which predominate in the, early stages of religious devf?lopment. Hinduism. We can pass on, therefore, to the next form of creed, which, under the title of Hinduism, is returned by more than 72 per cent, of the population of India. The clumsy name is only justifiable by convention, and only definable by the same process, of successive exclusion as was used above with reference to the application of the term Animism. Primarily and historically, it is the antithesis of Islam, and thus includes all Indian forms of faith in which the uncompromising Unitarianism of the adherents of the Prophet detected signs of the worship of idols. But for the present purpose greater discrimination is necessary, and internal dissent is entitled to recog- nition. For example, there are two purely Indian oft'shoots from the parent stock,, the systems adopted by the Sihhs and Jains respectively, that have been, on the whole, sufficiently differentiated to form distinct pi^ofessions, although, in the present day, the dividing lines are anything but clearly denoted, owing to the confusion bet-vyeen, social identity and doctrinal diversity. Buddhism, again, originally a manifestation of, nonconforinity arising almost in the centre of the orthodoxy of the day, received a, new set of characteristics when allowed to develop in partibus infidelium; and even in the few tracts within the borders of its native land where it still prevails, little affinity. ' is traceable between it and the religion in the bosom of which it was 'nurtured. There are two modern schools of doctrine to each of which the name of a separate religion is sometimes given, Brahmoism and Aryanism, but neither is current beyond a section of the community numerically insignificant, or differs sufficiently from the vital tenets of orthodoxy to be entitled to. such a distinction. In the return under review, accordingly, they are shown subordinately to the general title, and in the other returns in which religion is specially recognised, they are included in the latter. The confusion between Hinduism and the lower or tribal forms of religion has been sufficiently mentioned in the foregoing paragraph. And in some parts of the country, as, for instance, the Panjdb and Central India, there is a marked disinclination to apply the term Hindu to members of the lower and impure castes, such as sweepers and village menials. In the Panjdb, however, such returns have been added to the Hindu, but in Central India many have been classed as Animistic which are shown elsewhere as within the pale of orthodoxy. Thus, by the process of exclusion, we reach the conclusion that Hinduism is the large residuum that is not Sikh, or Jain, or Buddhist, oi* professedly Animistic, or included in one of the foreign religions, such as Islam, Mazdaism, Christianity, or Hebraism. Thus limited, a more applicable title for it would be Brahmanism, which connotes its two chief characteristics in the present day, the recognition of inherited social status and the authority of a hereditary sacerdotalism. It must be understood, however, that this use of the term does not imply that absolute supremacy of the priestly class which induced M. Barth, a leading authority on the subject, to restrict the term to the earlier stage of the faith before the Buddhistic schism, nor does it imply that the sacerdotalism in question is invariably that of this class, for we have instances where the office is undertaken bV those of other sections of the community, such as the Jangam, amongst the Lingaiat heil'etics, 159 and the Gosains of the quasi-orthodox converts of the Assam valley. The expansion of the creed consequent on the recrudescence of Brahmanic influence that succeeded the decay of Buddhism in India has induced the authority just cited to employ the term Hinduism for the present system, but as the latter is the work of the Brahman, and the entire tendency, whether of orthodoxy or dissent, is in the same direction as during the earlier development, it seems inadvisable to ignore in the nomenclature the main factor in the actual conditions. In few, if any, other parts of the world can be found so apt an illustration of what Gribbon puts as an obiter ' dictvm into the mouth of the Roman magistrate, that in every country the form of superstition which has received the sanction of time and experience is the best adapted to the climate and its inhabitant's. Religion, in the etymological sense of the word, it is not, and never was. The binding element is only educed by active opposition on the part of some other form of faith, sufti as Islam, and is generally confined to that part of the country where the two communities are separated by race as well as^doctrine. Elsewhere, if there be any general feeling at all, it is limited to sectarian disputes, subordinate to social considerations. Its extra- ordinary vitality is due, obviously, to its power of adapting itself to many new circumstances, to the elasticity that accompanies the absence of dogma in the faith taught to the masses, and to the relegation of ceremonial and liturgy to social observances, and equally concrete worship. To quote Gribbon, again, — " The superstition of the people was not embittered by any mixture of theological rancour, nor was it confined by the chain of any speculative system. . . . The deities of a thousand groves sjud a thousand streams possessed in peace their local and respective influences. . . . Such was the mild spirit of antiquity that the nations were less attentive te the differences thin to the resemblances of their religious worship." This applies very fairly to the state of feeling in India in the present day. The faith of the masses is purely local, and Brahmanism, as a whole, is absolutely restricted to the Indian continent, from the Himalaya to Rdmnad, and from the Indus to the borders of Arrakan. Here it lies, as has been said, like a lake, in gentle motion within its banks, from local ferment below, or as its surface catches some stray gust of fresh ideas. Every now and then an island of Animism is sapped and crumbles into the depths till no trace of it remains, but beyond this there is no accretion affecting the dead level of the contents. The process by which the community has reached this stage of, so to speak, marking time in its religious development, is highly instructive, for, as its recent historian, M. Barth, writes : — "Nowhere else do we meet with circumstances, on the whole, so favourable for the study of the successive transformations and destiny, so to speak, of a polytheistic idea of the universe. Among all t;he kindred conceptions that we meet with, there is not another which has shown itself so vigorous, so flexible, so apt as this to assume the most diverse forms, and so dextrous in reconciling all extremes, from the most refined idealism to tlie grossest idolatry ; none has suceeded so well jn repairing its losses ; no one has possessed in such a high degree the power of producing and reproducing new sects, even great religions, and of resistinj;, by perpetual re-genesis in this way from itself, all the causes that might destroy it, at once those due to internal waste and those due to external opposition." Of the religion brought by the early Arya into the plains of India, little, if any trace is left, By comparison of the more archaic conceptions of the Mazdaistic creed with those of the same class in the Vedic, it seems that the worship of the elements was one of the chief features in both. But the earliest utterances on the subject in the collections that have come down to us are clearly stamped with the mark of their sacerdotal origin, and can only be accepted as the productions of an age far more advanced in religious conception than that in which the Arya crossed the mountains, leaving their nearest kin in possession of the field in Central Asia. In the Vedic literature there are invocations, no doubt, parts of which are far older than the rest, which were pieced together after the sacrifice with which they were connected had ffrown into the appanage of a hieratic body, and from these some notion of the Vedic theology may be obtained ; but this is very different from what must have preceded it, and between the two is drawn a curtain which bears no sign or representation of what it conceals Modern investigators, however, are satisfied that alongside of the higher Animistic ideas that are shadowed forth in the Vedic, or sacrificial literature, there must have been an exoteric, and lower, stratum of beliefs, of much the same sort as are found amongst undeveloped communities in other parts of the world. As the immigrants advanced eastward, this lower influence asserted itself, and the superstition of the dark races came into contact with that of the foreigner. The sacrifice, as a means of securing tl^© ni^terial fq,vour^ of the gods, had by this time become t^e naaip feature U4 '■■ 160 in the rpligion of the Arya, and it is curious to note that even in this early period tlje religious element was egregiously re&tricted, for the altar of each clan or family hadjby special prescription, to be erected as far as possible from that of any other member, of the community. This is, perhaps, only what is to be expected from the very, practical bargain which the sacrifice was the method of concluding with the Supernatural powers,, the benefits of which could not possibly^ be shared by any besides the ■ provider of. the, oblation. From the time when the sacrifice passed from being an office undertaken by several of the leading classes of thecommunity into a mystery in. the hands of a special body, the liturgy naturally ' closed lupon what it had already acqwjfpd, p.pd the production of fresh invocations ceased, lest the monopoly of the sacrificing priest should be endangered. Thus a close body of literature was created, and the text grew in importance over the spirit, as the conceptions of the community were popularised' aiway from their oAginal tradition, and all participation in communion with the supernatural, was withdrawn from the bulk of the .population. The effects pn the language of thg. Arya has been already mentioned. On the hieratic body the results were more direct. The whole duty of the Brahman was held to lie in the preservation of the verbal accuracy of the ancient invocations!, and in the interpretation thereof as the original tongue fell more and more into disuse-. With the, exception of the metempiri^al. speculations that began to be rife soon after the class had outgrown the duties, imposed on them in connection with ministration, the whole course ^of study was directed to the ritual, in which their male progeny were trained by long and arduous instruction. Women and the lower orders were excluded from this course, and, in fact, no class but that of the priests was admitted to what was at the time the only learning. Ere long, too, the Brdhman abandoned the practice of the sacrifice in favour of study and devotion, with a view to his own personal salvation. The hereditary sub-division of society was also developed so as to ensurethe permanent, debasement of the masses of the people, as seems to have been the intention of the later authorities amongst the Jews, in the Levitical stage of their history. But the impassable gulf between the hierarchy sjnd the rest, the restriction to the fornaer of all knowledge, study and ritual, still more, the selfishness of their system, under which all effort was centred in the thaumaturgical, storage of merit for personal use only, to; the exclusion of others — all these are traits that inevitably lead to a reaction. ; , Buddhism. This came in the shape of Buddhism. Regarding the question whether such a person as Sakya Muni ever existed, there is this to be said, that there certainly was a personal head to the movement, and that to this fact is due a great deal, in fact, most, of the success that attended the propagation of the new creed in later years. On the other hand, there is no doubt that, as pointed out by M. Senart, most of the legend that attached itself to this personality in the hands of the proselytes who were not with the Founder from the beginning, is little more than can be attributed to. the ordinary glamour of a solar myth. But the main point to be noted in reference to our present subject is that the reaction grew out of the system itself, leaving much of the latter unafilected. The great change lay in the substitution of the merit of the good life for that of cult, theological lore, or asceticism. There was no overt oppo- sition offered to the current creed, but the flank of the priestly class was completely turned by the new doctrine. The monopoly of the knosrledge of the sacred works of Vedic literature by the Brdhman was not imp]igned, but Buddhism explained that such knowledge was not essential to emancipation. The practice of asceticism was not deprecated, provided that it was accompanied by active benevolence to others Worst of all, as a good life is within the capacity of all, the merit thereof could be reaped by anyone, irrespective of social positioUj so that salvation was placed within the reach of the lowest caste as much as within that of the Brahman or warrior. In the monastic discipline, which was described as the first step towards ;that object, caste distinctions, and afterwards those of sex, were obliterated. Human nature, however, was probably not so very different then from what it, is now, so there is no reason to conclude that this equality was at once carried to its extreme, save in theory • and as many Brahmans are reported to have joined the monasteries, it is more than probable that the impure castes, from long habit, held aloof till the new religion ha4 made considerable way. Just, so, in the present day, we find that in spite of the theoretical equality of all castes before the governing power, it has been found impossible to mix together the middle classes and the helots in the same school. The latter are either not admitted into the same building at all, or, Avhere they are provided with a separate quarter, the teacher is said to provide himself with a supply of clay pellets as instruments of chastisement, to avoid , the , poUutiqrt of the orthodox rattan 161 by contact with their bodies. But in this monastic discipline lay, in the first instance, the strength of the whole Buddhistic system, and its weakness in the end. The object was to train people to disseminate the principles of the faith by exhortation in the tongue of the masses. All was on a popular basis, and in the teaching there was nothing esoteric, but much morality and little or no 'spiritism. Buddhism made way slowly but steadily at first, without fanaticism or aggressiveness towards the Brahmanic tenets and practice. But, in accordance with the tendency, we have already had more than once to notice, the founder received the Iionour of apotheosis at the hands of hin disciples; miracles were attributed to him, and on the recognition of his system as the State church of the only consolidated monarchy in India at the time, the new faith enjoyed for rather over a century unprecedented oppoi'tunities of expansion. There was evidently opposition of some sort from without, and schismatics and Laodiceans crept into' the fold, but all this, so long as the path lay upward, only served to strengthen the faithful, and dissent was subdued by a General Council held under royal auspices. Selected missionaries were sent abroad with instructions from the king himself to see that the doctrine was made available to one and all in their respective vernacular languages. This last provision, however, was apparently dis- regarded in later years. But at all events Buddhism was conveyed across the Vindhya through the Deccan and the south to Ceylon. The mission to that island had the honour of being conducted by the son of the king in person, and from thence it spread later on to Burma, Java, and the surrounding country. The advent to power in the north of monarchs of considerable strength, both Skythian and Greek, gave an opportunity of diffusing the creed over the trans-Himalayan tracts bordering on India, and from thence into the north-east. Menander himself is claimed by the Buddhists as a convert, and there is no doubt that the human element in the creed carried it where the narrow doctrine of Brahmanism would receive no hearing. We have now to look briefly at the causes which led to the decline of Buddhism. We have seen that in the first instance it was the founder's intention to run parallel with the orthodox faith, and even to retain caste, where conduct was in conformity with Buddhistic prescription. But there was the obvious denial of the right of the Brahman, as such, to assume superiority over the rest, and this was the thorn that rankled. Harmony was outwardly maintained, however, until the favour of the monarch began to divert the stream of charity, which, under the old system, to be meritorious had to flow into the coffers of the Brahmans, to whom a gift was personal, into the laps of the monks, who by their vows endowed their monastery with all the largess they might receive. The results, apart from the effect on the Brahmanical mind, were twofold. In the first place, the monastic orders rapidly increased and throve, so as to be able to provide the army of trained missionaries of whom we spoke above,' and again, the establishment of these large bodies of more or less educated men in fixed situations, gave rise to architectural art in India, to the enrichment of the simple ritual of the founder, and, finally, to the erection of gorgeous temples and shrines over his relics. For, following on his canonisation came, of course, the desire to perpetuate his miraculous powers by the distribution of portions of his clothing, his begging bowl, and even his hair, teeth, and other terminals. But the tide of Buddhism as professed in India did not remain long at the full. We have no trust- worthy or complete account of the causes that led to. its decadence. It is plain, however, that soon after its accession to power as a State church it began to temper propaganda with persecution. One of the passages in the inscriptions of King Asoka puts this point with Thukydidean terseness :— " In less than two years the gods that were worshipped as true in Jambudvipa have been rendered false, the result, not of my glory, but of my ?eal." Then too, in the monastic system it was forging the means of its own extinction. The number of monks increased till they formed a society of their own, apart from worldly interests and from the lower sympathies of the community. When the zeal of ocnversion and exhortation had spent itself, the monks had nothmg left to do but what is their fate in all countries where they exist, to withdraw from active life, and pass their time in discipline, introspection, and what Bacon calls vermiculate discussion. Thus their hold over the people and over their creed as a working religion, weakened. The Brahmans, during their period of depression, had laid to heart the methods of their adversaries Theological teachers arose who enrolled bodies of disciples, m rivalry of the monasteries of Buddhism. Their tenets were popularised, their ritual enriched and modified temnles were built with public and gorgeous ceremonial, and, more important than all this, the Brahmanic pantheon was avowedly opened to the most / V8388. X 162 widely popular of the local deities, male and female. Thus the regenerated a^d transformed Brahmanism was very soon in a position to .gain an easy victory Qver their enervated rivals. The rapidity and completeness of the decay of the faith, of Buddha, in spite of the short and partial revival undeir Shiladitya, is one of the most remarkable features in the history of Indian religion. Not that it has npt left a praiseworthy mark on the latter-day Brahmanism that succeeded it, for no doubt t}xe charity and tenderheartedness with the disregard of sacrifice that characterise the latter, are indirectly due to the lessons learnt by the masses whilst under Buddhistic influences, as they are alien to both the older Brahmanism and the Animism with which it is mingled. But as a religion, it has completely disappeared from the continent. We find the north-eastern portion of Kashmir and of the P^^rjab, the Lepcha and Bh6tia of the eastern Himalaya, and a few tribes, in eastern A-ssam still acknowledging the creed of Buddha in its northern, developmiqnt. ,In Nipal, too, it holds the higher valleys, though the lower have succumbed to the social attractiveness of Brahmanism. Omitting this last-named tract, the Buddhists of India number, ^ut, 243,000, of whom, again, one-half are found in the border land of Burma and, Arrakah. When we take into consideration, moreover, the number of Chinese traders- and of Burmese convicts in Indian prisons, the actual Buddhist population becomes very small indeed, and of the Aryan community professing this faith not a single representative remains. All that are not Burmese are of Thibetan extraction. , , . Jainism. A second olFshoot from the earlier Brahmanism is found in the Jain, a form of belief that still subsists and flourishes in India to this day. Its origin is veiled from us, but it bears a strong family likeness to the earlier form of Buddhism, and it is a question amongst scholars whether it rose about the same time or a little earlier. At all events it seems to have been unpopular with the Buddhists, and to have diverged less from Brahmanic orthodoxy. The monastic system was not countenanced, but ritual was simplified and women were allowed to share in it. As in Buddhism, however, the larger section of the Jains decline to allow that women can attain Nirvana. The latter, however, is , with them perpetual bliss, instead of complete, annihilation. Caste, amongst the Jains, is maintained, and though they have no special reservation of the priesthood to a class, there is a general tendency in that direction, and in some cases Brahmans even are ernployed. In later years the Jains seem , to have competed with the Brahmans in literature and science, so that they fell into disfavour, and would very probably have succumbed, but for the advent of the Musalmdn power. In the north and west of India they are still a cultivated class, mostly engaged in commerce, whilst in the south, where they share with the Buddhists, who. preceded them, the credit of forming the Kanarese and Tamil literature, they are as a rule, agriculturists. Except in a few of the larger cities of the north there seems to be little sectarian hostility between them and the orthodox ; and in the west, where, they are still closer in customs and observances, the line of division is scarcely trace- able. In parts of both tracts there is, in the present day, a tendency for Jainism to regard itself as a sect of Brahmanism, in spite of the non-recognition of the divine authority, of the Veda. It is probable that in compliance with this tendency majUy have returned their religion as Hindu of the Jain sect, so that where sect is" not separately compiled, as in the Imperial series of returns, the total of the Jain religion is reduced by that number. As it is, the number of Jains is given m about 1,417,000. The Neo- To describe in other than very general terms the religion that supersede^ the Brahmanism. Buddhism of about the first century before and after the Christian era is a, task' that would require a volume to itself of very considerable, size. It will be sufficient to give a mere sketch of its more prominent characteristics. In the first place, the Brahmans, in whose hands the whole of the reconstruction rested, had to reconcile the popular element implanted by Buddha's teaching with their own supremacy. The first method adopted was the incorporation into their system of the local gods and goddesses, which seem never to have been ignored by the masses, whatever may have been the vi^ws taken of their power by the educated portion of the population. These divinities were accordingly proved to belong to the Vedic pantheon. The Veda was held up as the key of all knowledge, and the knowledge of its prescriptions alone gave man the power of controlling the supernatural. These notions were adapted to the popular under- standing by means of the later Brahmanic literature, epic and Pur4nic. Bv the term used here it is not, of course, meant that the authors were all of the Brahmanic east.e, for it is' well knoy^n that the most celebrated works of this class are attributed to others, 'but 'it is essential that all should bear testimony to 'the authority-' of the Veda. 163 Now the knowledge and interpretation of the latter was the monopoly of the Brahman c&ste, and we have seen that this fact was not disputed" by even the most strenuous opponents bf .the orthodox system as a whole. It is immaterial, accordingly, to "what number of deities the pantheon is opened, so long as they derive their recognition from, and are alone approachable through, the Veda. In the actual ceremonial of worship the Br-ahman takes no part, according to strict interpretation of the later Vedic prescription, and those that are occupied about a temple, either as ininistrants, or as guides to pilgrims, administrators of bathing-places, or other offices of a like character, are regarded as of a very inferior type of Brdhman, and scarcely recognised as of the caste at aU by the higher grades of the hierarchy. In fact, there are many functions which in the eyes of public Opinion are not acceptable to the divinity, if undeirtaken by a Brahman, another testimony to the noli- Aryan origin of the present theology. But the authority of the Veda is found in all social ceremonies or those unconnected with worship. Here, the sacred text is all important, and only a Brahman is compatent to enunciate it.' That is, it has become the spell by which the local deities are influenced, and public opinion, which is that of the caste, demands that the precautionary measure implied by its use shall not be disregarded. It has fallen a victim, in fact, to the particularism of which religion in India' furnishes so many exandples. Caste has been called the stronghold of the Brahmanic religion, in spite of its apparently purely social attributes, and it is to caste uses that the Veda has been inextricably bound. Sect, again, of which there is so much heard in the accounts of the Neo-Brahmanic system, begins, where it constitutes a radical reformation, in declaring itself open to all castes. Very shortly afterwards, it finds it has to except the leather- worker and the watchman, and, if it does not become a shadowy figment of doctrinal hair-splitting, turns intp a caste, with all its exclusiveness, and its tendency . to sub-division, as its number waxes. Next to caste in their hold on the minds of the masses, and it is of them alone, not the educated. few, that we speak, come pilgrimage and the deification of the teacher, or spiritual guide, both of which owe their origin, in all probability, to the doctrine of faith current in this later form of religion. Into this it is out of place to enter, savq to point out that it is itself a derivative, apparently, of the direction of worship to a special protective divinity. From the general conception,, it was very soon detached ijn favour of the concrete figure of an incarnation, and from the adoration of ' the la]l^t^r afc the mosfc efficient shrines,, it . is a short step to that of the jnortal to whom is due all the knowledge one ; possesses of that worship, more especially as the age of miracles is by no means past and the canonised- teacher is as competent to perform them as. any god in the pantheon. As an act of faith is in itself an absolution, pilgrimages are the substitutes for most other religious observances,, always excepting those connected with the propitiation of local spirits, whether family or attached to the house or village. It may be worth remarking, too, that in south India, yrhere some local form of belief must have attained considerable develo-oment before the Arya i^nfluence on it began to be felt, and perhaps in Lower Bengal, also, for the same reason, the number of local deities, especially those requiring to be attended by people, of the helot class, is much larger than in Upper India,, and the propitiation of the malevolent spirits holds a far more important place in the popular ceremonial. The Br9,hnianioal faith, too, was more developed when ifc reached the south, and- as its greatest influence was probably exercised after it ha_d begun its reconstruction on the new plan, the Dravidian element that it had to assimilate was proporti6.n;ately high. The decline of a polytheistic religion, too, it has been said, is traceable in the exaltation of female deities, as well as in the multiplication of those of the other sex, and in both Madras and Bengal we find this feature very prominent. It almost seems, moreover, that in India, at all events, the weaker the race, the more numerous and more bloodtliirsty are its gods, and the greater the. influence of the sacerdotal system of caste. In connection with the present development of Brahmanism there is no need to speak of the contact with it of foreign religions, since whatever its eflFect on them may have been, they have made no impression on it, though the progress of conversion in one or two cases has been considerable, As regards Buddhism, the greatest, if not the earliest, of the ofi"shoots from Neo- Brahmanism there is but little to add to the historical summary above given, except in Buddliism. relation to local developments. In the Himalayan form of this creed, which received its shape from the daimonolatry qf the Skyl^hian races which it replaced, the original tenets' have been quite overlaid by the incorporation of former Animism. The Lama has befcome a priest and exorcist, instead of being a monk and preacher, and where the Indian f^^hti^r is apprdached the caste system of ' modern Brahmanism has attacked X 2 ]C"4 the tribal communities of the Buddhists. In one valley, fur instance, the census shows that the inhabitants are in doubt as to what their religion may be, for in 1881 they were all set down as Hindus, and on this occasion they appear as Buddhistsi In the eastern Himalaya the Lamaism is more restrictive, and of Mp41 there is no occasion to speak. In Assam there has always been a suspicion of serpent worship and other forms of Animism, and the Buddhism of the days of the Chinese pilgrim bore but a poor character, as the inhabitants were, he says, addicted to the worship of Deva, or local gods, and to sacrifices. In the present day the Buddhism is confined to those who have apparently brought it with them from the north-east, so that it was altogether obliterated for some centuries, even if it ever had a firm hold of the wild tribes of the valley. ' The only country, therefore, of those with which the census has to deal, in which Buddhism still flourishes, is Burma, where it is the religion of about 90 per cent, of the population. But even here, as Mr. Eales points out in his report on the census, the popular belief is little but Animism, and the attachment to the higher creed is largely due to the educational influence of the religious orders, because every boy has to be sent, if only for a short period, to one of the monasteries as a novice or lay brother. Like, the Neo-Brahmanism, too, it is gradually absorbing within its sphere the forest tribes, who are professedly Animistic in their belief, and, like its former rival, it places no embargo on their tutelary gods, whilst as above stated, through the monastic system, it tends to raise them in the social scale. Considerably more than half the males in Burma can read and write, a feature in which the difference between the two systems is very clearly denoted. Sikhisni. Of Jaiuism as a present-day religion we have said enough above. Sikhism was not specially mentioned in the general sketch of the Brahmanism from which it is derived, as it is a comparatively modern movement, and one that is not connected with the general history of the later development of its parent, since it has been confined to a single province and almost to a single community. It originated towards the end of the 15th century, and seems to have been due to the teaching of one of the most influential of the sectarian leaders of a quasi-TJnitarian revival amongst the lower classes of Brahmanism. Kabir, the leader in question, was very appreciative of the tenets of Islam, on which ground he was alleged to belong to that faith, but in all other respects his doctrine and practice was that of Brahmanism, and there is no doubt that though he vnshed to combine part of Islam with the teaching he founded on the Veda, his views of the Kuran were anything but orthodox. The founder of the Sikhs, or disciples, Nanak by name, seems to have rejected both Kuran and Veda, but maintains Brahmanio practice in social matters, and still employs that caste in ceremonial. The influence of IsMm, however, as was natural to one of the frontier province, is traceable not only in his doctrine, but in the conception of the virtue of aggressive hostility towards dissentients, which was not brought to its full significance until the fifth Gruru from the founder espoused the side of the emperor's son in one of the numerous family disputes amongst the Moghals. He was promptly thrown into prison and died there. From that date the Sikhs became a political force, and have never since relaxed their hostility towards the Musalman. The conversion was completed by the 10th Guru, Govind Singh, who died early in the 18th century, and under his guidance the separation of the disciples from the Brahmanio community was effected. But the new departure seems to have been largely nominal only, and a good deal of orthodox practice remained intact. The political objects of the Sikh leaders obscured the doctrinal, and culminated in the establishment of the kingdom of Ranjit Singh. In the present day, peace has relaxed the bonds of discipline, and the distinction between Sikhs and the rest of the Brahmanio community is mainly ritualistic. For example, it was found by experience that at the census, the only trustworthy method of distinguishing this creed was to ask if the person in question repudiated the services of the barber and the tobacconist, for the precepts most strictly enforced nowadays are that the hair of the head and face must never be cut, and that smoking is a habit to be absolutely avoided. The observance of these two injunctions, writes Mr. Maclagan, the Census Superintendent for the Panjab, is the best practical test of true Sikhism, " without any further rigmarole about sects " ; for not only is a true Sikh generally called a Hindu in common parlance, but many of those who are spoken of as Sikha are not true Sikhs, but Hindus. It is this confusion between the sectarian and the ceremonial elements that makes the return of Sikhs liable to correction, and instead of the number shown in Table VI. of the Imperial series, which is 1,907,833, the Panjab Superintendent estimates that, so far as his Province is coDcerned, which is the only one in which this religion prevails to any significant 165 degree, about 86,000 should be added from tlie Brahmanic totals and that 488,000 should be deducted from the Sikhs and added to the latter. The first item, he adds, is fairly accurate, but of the so-called Ndnakpanthi, or sectarian Sikhs, it is possible that some may be true Sikhs as well. It is as well t6 complete this review of the Brahmanic and quasi-Brahmanic Modem religions by adding a few lines on the two modern developments of Br^hmoism and reforms. Aryanism. The former is more or less known in Europe from its adoption by the scholar Ram Mohan Eai, who came to England, and dieid there in 1833. It originated in Bengal, and whatever stimulus it has since received has been due to teachers belonging to that Province. Some 30 yeiars after its foundation the usual detach- ment took place of a party that desired greater communion with the orthodox Brahmanism from which they started, and the younger spirits that proposed inde- pendence, based on the eclectic assimilation of what they thought applicable from other religions. The leader of the innovators lost his influence, in great measure, by his approval of the marriage of his infant daughter with Brahmanic rites to a young ^arti of considerable mundane attractions, and bn^is death shortly afterwards, another split took place, in consequence of the disputed succession to the post of leader. The movement is purely local, and confined to the literate classes of Bengal Hindus, with a few converts, it is said, from IsMm. The total number of those who returned this form of faith, either as their religion or as a sect of Brahmanism, is 3,051, out of whom 2,596 are in Bengal, 239 in Assam, and 128 Bengalis serving in Government or railway employ in the Panjab. The sporadic entries under this head in other parts of India are generally debi table to the somewhat similar sect of the Aryas, which is noticed below. As regards the number of the Brahmoists, it is highly probable that there are many more sympathisers with the doctrine than are shown in the census return, but, owing to the natural dislike to break from their social moorings in the haven of orthodoxy, a result which might follow on open sympathy, they adhere to the original term, and hold their creed to consist of a sectarian distinction, which commits them to nothing when marriage or caste ceremonies are in question. The Aryan schism is based on much the same lines as the Brahmo as regards its relation to Brahmanic orthodoxy. Doctrinally, it seems to be something in the nature of the " Veda up to date," and finds that these sacred works contain all the discoveries of modern sciences, including electricity and natural selection. In theory, the study of the Yeda is opened by them to all, but in practice, caste is maintained in all its strictness, and, as most of the adherents of the Neo-Arya faith are of the Writing or Trading class, a line is drawn below these, but none between them and the Brahman. The census return shows the number of this sect to be just under 40,000,. It was founded about 1877, in Lahore, by a Brahman from western India, who retired to Aimer, where he died in 1883, some say by poison. Most of its adherents are found in the Panjab, North-West Provinces, and the surrounding tracts in which the clerical agency required in State or railway offices is recruited from those provinces. It is interesting, as a study of the results of western instruction on a hereditary literary body in social subordination to an equally inaccessible hierarchy. It makes no appeal to the masses, and as the female element in the return is mostly illiterate, and the founder has not yet been canonised, it does not seem likely to supersede orthodoxy at the family hearth, and will probably manifest no more vitality or influence than any bther of the almost innumerable sects of the Neo-Brahmanism. It has achieved the first step towards success in the possession of an adversary, a " religion " started by a personal enemy of its founder. The new faith was proclaimed on the occasion of the Jubilee of Her Majesty, at Lahore, and is now estimated to haye 190 followers. It publishes three periodicals and a shower of tracts, mostly directed against its rival. The latter is not slow to retort, and as both parties are recruited chiefly from the officio-clerio community, to whom composition is a recreation, the amount of literature produced, in proportion to the number of the faithful, is enormous. We have now to consider the religions foreign to India. First of these may be Mazdaism placed that of the Parsis, entitled Mazdaism, from the name of its supreme deity, or (^«^''«)- Zbroastrianism, from the Greek rendering of Zarathlistra, the reputed founder of the creed Its claim to be treated of next after the religions that have taken their rise m India rests upon the common origin of Mazdaism and Brahmanism. In both cases we find the same gods named, but as the community split into two branches before the rrfision Had b6ep ^t all organised, it is impossible to trace more than a vague outline, ^is tiBttal in such cases of early dissent, the gods ot the one branch became the devils X 3 166 of tHe other, but whetlier, as some think, tte Indian branch went south and kept on attacking the Eranic branch ■ from their standpoint on the upper Indus, or whether the two quarrelled on the north side of the mountains, and separated in consequence, seems an open question, and likely to remain so. -In the reign of the Persian monarch, Darius the First, the two resumed contact, as the former obtained possession of the north-west corner of India. But his hold was slight, and the in|;ercourse, even if not barred by the intervention of, more than, one chief of Sky^hian race, was barren of results. By this time, moreover-, the religion of the Eranic branch had been moulded in quite a different form from that taken by the early Arya of India, It is anqther open question whether the founder of the former, Zarathiistra, had any more real existence than, Buddha, but it is certain that in the time of Parius, the religion attributed to his agency was established as that of the State, throughout Persia, ^he invocationary literature, as with that in the Yeda, consists clearly of compositions of very different age. Most of it bears the trace of the influence of ■ the forms of belief that prevailed at the time in Mesopotamia, that great birthplace of creeds, under which influence it diverged completely from the luxuriant growth taken by the Aryan mind in the more genial conditions of , the south. The, Greek invasion drove the Mazdian religion into temporary obscurity, but during the rule of the dynasty of Sasan it achieved a temporary revival, and to this period is due the greater part of the Mazdaistic liturgy, which is in the Pahlavi language, in contradistmction to that of the Avasta, or original scripture. Then came the flood of Islam, which directed all its force to the extirpation of what it denominated fire-worship, with such efficiency that only in the north-west corner of what is now Bran was any remnant left of the Eranic faith.* This , brings the history of that faith down to the point with which we are chiefly concerned. A portion of those who preferred exile to either the abandonment of their creed or massacre, left their native country by the Persian Grulf for Ormuz, where they resided for some years. In the year 717 they reached India, debarking at Sanjdn, a small place some 60 miles north of Bombay. Here they settled, and four years later built their first place of worship in the land of their adoption. It would be pleasant to be able to attach thorough credence to the legend that with them they brought the sacred fire from their home in EraUj as Mneas conveyed his household gods from the flames of Troy to their new home In Latium. For several centuries the Parsis, as we may call them in their new domicile, peacefully waxed and multiplied. On several occasions they rendered important services to the local chief, who requited them with grants of land for settlements. Some of their number even visited the court of the Moghal Emperor, Akbar, who favoured Mazdaism with the affectation of temporary adhesion. D'p to the middle or end of the 18th century, Surat and Nausari ifvere their headquarters, with colonies all over Grujarath and even in more distant parts of India. They have ever since that time been gradually transferring themselves and their interests to the city of Bombay and its suburbs, on account of the opening they found there under British auspices for their great enterprise and commercial talents. Their religion during the early years of their residence in India suffered from the influence which must be exercised over a small and foreign body, numerically weak, by a strong Animistic atmosphere, combined with that absorptive power which is, as we have seen, one of the most prominent characteristics' of the Brahmanism of the day ; and from the accounts of coiitemporary travellers it appears that at one time it was barely distinguishable froni the surrounding form of worship. But reinforcements arrived from Eran, and race-feeling has siiice served to keep it unmodified in all inain doctrine, so that the lower influence is iibw confined to social and semi-religious ceremonial. On the other hand, the , long period of political extinction in such a small community, jealous amongst foreigners of the tenets of the faith they had taken such pains to preserve, checked the development of their creed beyond the limits of the scripture and ordinances of rite and ceremony to which thev attached so patriotic a value. In the present day the Parsi is a completely Indian community. The language they brought with them from Eran has been relegated to their liturgy, where it is studied like a classic by the scholars of their race. The mass of the Parsis in speaking and writing use Grujarathi, the vernacular tongue of the tract where they first landed, and which is said to have been imposed on them as a condition of residence. In 6ertain circles, especially in Bombay, the census shows a considerable tendency to return English as their mother tongue, a statement which * So!ae writers have said that the Eranic religion was destroyed by the Turk invaders afld.that tliose' ' d '^ Abu-Bakar alUowed its exercise in-peace, but there must have been- strong grounds for ihe Voluntary p■x'^'k• ^f so small rt, band- and the cessation -of conimunication^on their part with their native oountry. . ; 167 will be by no means impeobable in a few years, wben the girls of tbe present generation arrive at _the dignity of maternity , ior tMs race has shown- a remarkable, aptitude for assimilating the lighter tints of western life, so that it will be interesting to those of a few igenerations , hence to see in what way this occidentalisation has affected the religious development of the community. Taking the whole of India together, the Parsis number Just under 90,000, but on looking into some of the details, it -seems likely that a few Persians may have been here included, owing to the similarity of the names in some of the vernacular characters. The groimds for this supposition is that there are a few Parsi entries under the Musalmdn religion, and if there are any twq creeds in, India the yptaries. of which are absolutely antagppistic to each other, they are Islam and that whiqh it, did its best to destroy. On the borders of the Caspian and a little to the south, in Yez^,, there are still some of the ancient fire temples, with a congregation epima^ed at frqm 6,000 to, 10,000 souls, with, whom, during the last 20 yeai;s or so, the Parsi^ of Bombay have re-opened communications, so that when a famine "took place in 1871, or thereabouts, a considerable number were brought down from Yezd to Bombay, where they were maintained by their more fortunate co-religionists. Of the Jews and their religion little need be said here. Apart from the modern Judaism, immigration to the chief commercial cities, such as Bombay and Calcutta, there are two colonjles of this race, both on the wiest, coast. It seems probable that both arrived by sea, and though the date of their anabasis is not known, there is no doubt that they are of very considerable antiquity. In both cases are there two sections in the community ; one, called the white, the other that of the black. The former justify their title by the tradition that they have kept aloof from intermarriage with the daugh.ters of the land, whilst the others,, like the Arya, have fallen victims to native alliances. In addition to these, though, not strictly speaking within India, we find a, colony of ti;ader3 of this race in the settlement of Aden. These, being, nearer to, their native country, are of ;piTrer race and less itainted "with local custom in their religion. They speak, top, a northern dialect of Arabic. The whole number, including these last, is only aiDou,t 17,200, and of these 10,500 g.re found in Bombay, 2,800 in Aden, 1,300 in Cochin, and, 1,450 in Calcutta. The rest are scattered over India, both in trading centres, and, in, the case of the Bombay cpast colonists, in the ranks of the, , native army. These last, whilst maintaining the .main principles of their faith intact, have adopted the language and much of the social custom of the Maratha population by whom they are surrounded. It, is much the same with the Israelites of Cochin. One fifth of the population of India is returned as of the faith of Islam, but these Islam. 57 millions, like the 208 millions of Brahmanism, though to a less degree, contain an extraordinary collection of heterogeneous elements. In the first placd we must divide the community into those of foreign race, who brought their faith into the country along with them, and those who were converted from one of the previous creeds.' of the country itself under various inducements and at many different periods of history. Not counting the teachers of Islam, who are scattered pretty numerously over India, the foreign colony, speaking generally, consists of three main bodies ; first, and numerically the most important, are the immigrants from just across the north-west frontier, who are found chiefly in Sindh and the Panjab. Next come the descendants of the court and armies of the Musalman dynasties, who are also centred in upper India, with considerable offshoots in the large feudatory States of the Deccan ; .lastly, we have all along the west coast, settlements probably of Arab descent, who- arrived by sea. Some of these, especially those who took to agriculture, appear to have mixed with the natives of India, and to have adopted a good deal of the social system of the latter, whilst others, including the trading classes, who could keep touch more easily with their country of origin,, are decidedly of purer descent, and have ])reserved, moreover, their faith in much more of its primitive form. As regards the converts, no such subdivision is practicable. It is true that the process of conversion in India was of two very different characters, only one of which, is still extant; but at the present time we cannot, distinguish, at least in the parts of the country where both have prevailed on different occasions, the results of > the. one from those of the other. It seems probable that the first in the field was the method of peace introduced by those from over ^ea. , Omitting considerations of doctrine, the new faith held out inducements of great worldly attractiveness, especially to those whom the Bra;hmanic regime had chained to a position of perpetual subordination. We find this along the west eoast, and it is in full: operation in the eastern division of Bengal at this moment. The official propagation of Islam does not seem to have been undertaken' far < several X 4 168 reigns after the establishment of the Musalman power in northern India. The eai'ly monarchs, though no doubt they took no steps to quench the iconoclastic ardour of their followers, had their attention fully occupied by the novelty of their administrative task ; and it is also fairly plain that Babar and his family were not troubled with much anxiety for the salvation of their infidel subjects, since they were by no means convinced of the exclusive merit of the substitute they had to offer. It was not until the tolerance of Akbar and the indifference of his two immediate successors had given place to the zeal of Aurangzib that the diffusion of the faith of the Prophet was taken in hand with real fervour. From this time downwards we hear of wholesale conversions, some to avoid the alternative of death or confiscation of property, others to gain the favour of the great. The same system was pursued later on in the Deccan, when the Sultdn of Mysore was seized by the desira to add to the number of the faithful. It does not appear that more than the public profession of belief was required, with the requisite punctuality of attendance at the Friday service and the observance of the prescribed fasts, so that all that is of the more vital importance to the masses in India — especially to the women — was left as before. The simplicity of the Kuran made but a feeble appeal to the mind of those accustomed to the sensuous rites of the new Brahmanism and the neighbourly proximity of the supernatural in the Animism that governed the life of the people at large. Hence arose the remarkable differences that demarcate the IsMm of India from that of the population amongst whom the religion originated. In the former the " mild spirit of antiquity," as we have seen above, mitigated the asperity of sectarian hostility, so that Shiah and Stinni joined issue in doctrinal discussion without recourse to arms. The good men amongst the teachers received divine honours as if they had never left the Brahmanic fold, and in default of the pilgrimage to Mecca, which was beyond the reach of the majority, resort was had to the tombs of the canonised, where fruit and flowers are offered, as to one of the orthodox pantheon and often by Hindu and Musalman alike. Saints are the special feature of the Indian development of IsMm, and the worship of relics follows. In some places there is a hair or two, in others a slipper, elsewhere a footprint, of the Prophet, to which the devout pay homage and are rewarded by miracles. Even where the two religions do not participate in the same festival, the more simple has borrowed for Indian use some of the attributes of the more elaborate, as in the case of the procession of the tombs at the Muharram, and the subsequent dipping of the imitation fabrics in water, as in the Durga Puja of Bengal. In many instances, where the two forms of faith exist more or less in numerical equality side by side, the Brahman officiates at all family ceremonial, and, as it has been put by a local writer, the convert to Islam observes the feasts of both religions and the fasts of neither. This state of thought is very, much like that described by Lady M. Wortley Montague as existing amongst the Makedonian Arnauts of her time, who, living between Christians and Mahomedans, and — "Not being skilled in controversy, dcclaiu that they are utterly unable to judge which religion is the best, but to be certain of not rejecting the trutn, they very prudently follow both. They go to the mosque on Friday and to the church on Sunday, saying for their excuse that at the day of judgment, they are sure of protection from the irue Prophet, but which that is, they are not able to detcimine in this world." Compare with this the following experience in the present generation as having occurred in the eastern plains of the Panjdb : — '• A traveller entering a rest-house in a Musalman village found the headman refreshing the idol with a new coat of oil whilst a Brahman read holy texts alongside. The pair seemed rather ashamed at beino- caught in the act ; but, on being pressed, explained that their Mulla (priest) had lately visited them and being extremely angry on seeing the idol, had made them bury it in the sand. But now the Mnlla had gone, they were afraid of the possible consequences, and ^vere endeavouring to console the god ior his rough treatment." On the frontier of this province, where the bulk of the population is Musalmdn, and fanatically devoted to that faith, the Animism takes a different shape, and the Miilla is simply a magician. In the southern tract the spiritual guide is valued it is said, by his supposed ability to procure the object of the vows of his disciples. In the north, the description given by Sir Herbert Bdwardes is so graphic that its reproduction will be excused : — " For the Bannuchi peasant the whistle of the far thrown bullet, or the nearer sheen of his enemy's sword, had no terrors ; blood was simply a red lluid ; and to remove a neighbour's head at the shoulder as easy as cutting cucumbers. But to be cursed in Arabic, or in anything that sounded like it to be 169 told that the blessed Prophet had put a black mark against his soul for not giving his best field to ou^ of the Prophet's, own posterity ;i to have the saliva of a disappointed saint left in anger on his door-post, or b,ehold a Haji who had gone three times to Mecca, deliberately sit down and enchant his camels with the itch, and his sheep wifih the rot ; these were things that made the dagger drop out of the hand of the awe-stricken savage^ his knees to knock together, his liver turn to water, and his parched tongue to be scarcely able to articulate a full and complete concession of the blasphemous demand." To find a parallel to ttis feeling, one has almost to go west of Bantry Bay. But the above remarks refer mainly to the rural population, and in the towns, where the masses of the Miisalman community are equally converts from Brdhmanism, the standard of doctrine and practice is better maintained. In the tracts, too, where conversion is still going on, and the bulk of the surrounding population is Brahmanic of the lower polydaimonic type, as in Bengal and the Malabar coast, the antagonism is sufficient to keep Islam alive. In the Panjab, on the other hand, the castp is weak, tribe and race are strong, the power of the Brahman is in abeyance, and religion is altogether on a popular basis, where supernaturalism is strictly localised, so as to be under the control of special persons or agencies. Conversion and reversion, therefore, amongst the lower classes are here common. "Where the caste system is very strict, on the other hand, the lower orders, instead of retaining their social titles on conversion, affiliate themselves to one of the foreign tribes. Shaikh, by preference, as has been mentioned in the introduction to this chapter, so that they obliterate the traces of their origin, and, if lucky in their worldly affairs, become the social equals of the best. One point regarding Islam in India may as well be noted here, as it is of considerable interest in connection with the general results of conversion ; it is that the prejudice against postponing tnarriage till the girl is grown up is very much weakened, even if not altogether obliterated, and this, as well as the absence of prohibition against widow-marriage, seems . to account, according to the views of Mr. O'Donnell, the Census Superintendent for Bengal, for the better lives of the Musalman population, and the consequent greater rapidity of its growth in that province. Something, too, may be due to the diet of the Musalmans, which in Bengal is more varied and nutritious than that of the Brahmanic castes, whilst in the Panjab, on the contrary, the restrictions on the Musalman are greater than those on his rivals, so that a year of high prices or scarcity is said to drive many of the lower class of the former into the ranks of Brahmanism, for the time being, to revert to the more liberal creed when its rules as to diet press less harshly on their resources. It is thus plain that in India the religion of the Prophet is in practice by no means the uncompromising Puritanism it is found to be in, say, Arabia, whatever may be the theoretical identity of the two forms. In the towns, indeed, antagonism between it and the local religions is always smouldering, and if the strong hand were to be lifted for a moment, there would be no loss of time in finding an excuse for collision. To the Musalman, Hindu, Sikh, and Parsi are alike obnoxious, as he to them, and against another religion, Sunni and Shiah fight shoulder to should'er, leaving their own sectarian differences to be settled after the disposal of the common enemy. Christianity was probably the latest of the great religions to be disseminated in Christianity. India. It is true that a small settlement of early Christians was established on the west coast within the first two centuries of that era,* but the real spread of the creed dates from the 16th century, when the Portuguese navigator, Oabral, was instructed by royal commission to conquer territory and to promote Christianity, beginning the latter task " with preaching, and proceeding, if that failed, to the sharp determination of the sword." The methods adopted were, in fact, political, and the support of the Portuguese power was the reward of profession of Christianity by any of the petty chiefs of the tracts surrounding the European settlements on the western coast. A generation later, with the arrival of ' St. Francis Xavier, missionary enterprise was not only extended, but took a milder shape, detached from the polemic zeal of the first Portuguese Crusaders. The missionaries assumed the habits, dress, and often the titles of Brahmanic ascetics. It being their avowed object to build on an indigenous foundation in order that the priesthood might be recruited from amongst the people themselves' they recognised' much of the local Brahmanic customs, such as that of * The arrival and death of St. Thomas on the east coast of Madras is probably apocryphal, as it seems that his labours terminated in the Skythian or Indo-Baktrian territories of Gondofares, apparently in lower Sindh ' '"ut there were Christians whose conversion was attributed to his influence settled on the Malabar coast at the end of the second century, who were visited by a Roman nayigator ajbout the year 190. I 78388. 170 caste. The upper orders were separated in church from the lower, and intermarriage between castes was not sanctioned. Then, again, the creed had its miracles and martyrs to hold up to their congregation, its relics and acts of faith. Pilgrimages are still made to the tomb of St. Francis Xavier in Groa, on which occasions devout aspirations are fulfilled and fleshly ills are cured. These manifestations are probably now confined to the faithful, but in old days, Hindu and Musalman joined at the tomb of Albuqueraue in imprecations against his successors. In Upper India, the mission of the Jesuits and others, though less fruitful in numerical results, were apparently subjected to less opposition from the local rulers, as was to be expected from the policy of Akbar and his son towards the creed of any denomination. In the present day, the missions of the Eoman Church are widely spread over India, but their chief strongholds in British territory are found in the extreme south and on the west coast of the peninsula. Recently, too, their propagandist zeal seems to have been remarkably successful amongst the dark tribes of the hills of western Benga,!. Missionary enterprise on the part of the reformed churches of Christianity was two centuries later in its inception. Its pioneers were Danish Lutherans, who began work in the tracts surrounding Tranquebar on the south east coast, just above the Roman field of labour on that side of India. At the end of the century, too, the Euglish mission to Bengal established itself under Danish protection, on the Hughli, and there carried out the literary work which has made the name of Serampur so famous. In 1813 the Established Church of England entered the field. A bishopric of Calcutta was established, with archdeaconries for the two other Presidencies. In the next year, missionaries were sent out, followed, 12 years later, by another society. Since then bishops and societies have greatly multiplied. Some of the latter spread themselves all over the coimtry, others settle down in selected localities, and confine their labours to it. We find amongst the nationalities represented, not only the principal ones recognised at the Sacred College, by whom the various ecclesiastical divisions of the Roman Church in India are respectively manned, but the Reformed Church comprises English, Welsh, Scotch, Irish, American, Swiss, German, Danish, Moravian, and Swedish. The various denominations represented are given in Supplementary Table A., at page 495 of the first volume of returns, for in the case of this religion, the tabulation of sects was prescribed as part of the Imperial scheme, instead of being left, as in that of the other religions, optional with the provincial authorities. Unfortunately, for spite of special instructions to secure accurate nomenclature,' the results show that for a considerable proportion of the Christian population, the denomination was not returned at all, or else entered under the general title of Protestant. Thanks, however, to tlie courtesy of the Secretary of the Church Missionary Society, reference has been made praciiicable to a volume of Statistical Tables of Protestant Missions for 1890, through the aid of which the mission to which the above indefinite entries should, from their locality, be probably assigned, 'lias been ascertained. Taking into consideration this adjustment, the return shows that out of a total of 2,284,380 Christians, 57 ^ per cent, belong to the Roman Church, 8f to the Jacobite or Syrian Church, 33 per cent, to the various denominations of the Reformed : Church. The rest comprises the unspecified, and a few members of the Greek and Armenian forms of creed. The details will be considered below, where the statistics of each religion are dealt with in more detail. Minor forms In every enumeration of forms of faith room has to be found for a residuum of religion, of special classes of belief which cannot be fitted into any of the corners of those more generally prevalent. In India this residuum is small but varied. It is not, however, complete, for we find samples of Unitarians, Theists, and Agnostics included with Christians, probably because the enumerator found them mostly in the schedules of Europeans. Similarly, there were a few instances, altered on scrutiny, of Buddhists amongst European and Eurasian Christians. Some of these related probably to the professors of the new Theosophy ; others may be set down, perhaps, to the example of a high ofiBcial, who pronounced for Buddhism, on the ground that there was less to be said against it, he thought, than against any other form of religion. The number of persons ft-om whom the return of religion was not required, or who did not fill in the column reserved for it, was remarkably small, so that no more than 42,578 people are excluded from the table. These are not to be confounded with those who positively assert that they have no religion, of whom there appear to be IS, including one whose attitude was further emphasised by the entry of ""indifferent." 171 Per-ceutage Approximate Religion. Population on Per-centage of (1891). Total Variation Population. since 1881. Biahraanic 207,731,727 72-33 "1 3-23 / + 10-82 Animistic 9,280,467 Sikli 1,907,833 0-66 -4- 2-12 Jain 1,416,638 0-49 + 14-86 Zoroastrian 89,904 0-03 + 4-91 Buddhist 7,131,361 2-48 + 24-46 Jew - - - 17,194 0-006 + 20-93 Christian 2,284,380 0-80 + 21-85 Musalman 57,321,164 19-96 + 10-61 Minor forms 185 Unreturned 42,578 0-014 — Total - 287,223,431 100 GO + 10-98 We have now to deal tvith the returns which are to be found in Table VI. on Statistics. page 87 of the first volume of the Imperial series, and in a summarised form on page vi. of the same volume. The three aspects in which they are to be regarded are their respective and local prevalence, and their variation, wherever it is possible to ascertain it, from the corresponding return in 1881. The marginal table summarises the first and last, and in the table on the next page will be seen the proportion borne by each to the total population of the Province or State. Wherever there are special features to be noted with regard to the territorial distribution, mention of them will be found in the text. Comparison between the two years is in some cases little better than approximately correct, as has been admitted in the heading of the marginal table, since to uncertainty , as to the exact limits of the tracts not enumerated by religion in 1881, there is added that regarding the identity of the definition and application of the nomenclature on the two occasions. On looking over the provincial details for the purpose of compiling this statement, the latter defect was found very prominent in the case of Sikhs, especially in Sindh, and of Jains, in most of those parts of the country where they are only found as exotics. The territories not brought under the census of 1881 have been, of course, excluded in making the comparison, but, in the case of Brahmanism, there may possibly be a little excess in the 1891 return, as compared with the one preceding it, owing to the inclusion of some parts of the wild tracts of Rajputana, Madras Agency, and Eastern Bengal, but looking at the numbers dealt with, the difference in the proportional figures will be insignificant. The mean proportion of the Brahmanic to the total population is 724 per cent., Territorial and the table under review shows that this ratio is exceeded in all parts of India, distribution except Burma, the Panjab, Kashmer, Assam, Bengal, and Sindh. The detached Settlements also show greater variety of creed, as is to.be expected, and the small State of Eampur, under the North-West Province Government, reduces the proportion in that group, owing to the preponderance of the followers of the faith of the ruling family. In the case of Kashmer, Sindh, and Bengal, too, the main competitor in popular favour is Islam, but in the Panjab and its States, the Sikh faith, and in Assam, tribal Animism, are considerable factors in the distribution. Burma is almost exclusively Buddhist, but it is interesting to note the difference between the prevalence of that creed where it has been for generations the State religion, as in the Upper division (96| percent.), and where non-sectarian administration prevails, as in the lower, and it has had to compete with the local Animism on the one side and Christian missionary enterprise on the other, not to mention a considerable admixture of Brahmanism casually introduced by the labouring population from Madras and Bengal, all of which reduce its proportion to 86 "8 per cent. It may be remarked, in passing, that in Kashmer a purely Brahmanic administration is placed over a pDpulation of which but 27 per cent, profess that creed, whilst, on the other hand, in Haidrabad, a strictly Musalman regime presides over the destinies of a community of which 90 per cent, are, in its eyes, doomed to perdition. Taking Brahmanism as returned in the tables, its greatest relative prevalence is found in the south of India, Mysore, Coorg, Madras and Haidrabad ; Bombay, Berar, and Baroda follow close behind, and eight "other Provinces or States, including Rajputana and the populous tracts of Oudh and the North- West, shows this creed in a proportion of over 80 per cent, of the total popu- lation. It has been sufficiently shown in the foregoing portion of this section that the identity of the title does not imply anything like identity of doctrine or practice, except in regard to caste and pilgrimage, and the variety is on the whole local rather than dogmatic. Certain sects attain popular acceptance throughout a certain class of the community, not always in the tract in which they originate, and are there localised. Y 2 of religions : Brahnianists, or Hindus. 172" Table showing Territorial Distribution of Population by iReligion^ ^ riT f I Province or State, &c. Proportion of each Religion to 10,000 of the Population of the Province or State. S ,q a a 'S i-s .a O ^ Ajmer Assam Bengal ,, States Barar Bombay States Sindh Aden Burma, Upper „ Lower CentralProvinces Coorg Madras „ States N.-W. Provinces States States Oudh Panjab „ States Quettah Andamaus Haidrabad Baroda Mysore Kashmer Rajputana Central India "r Provinces I States INDIA 8,075 5,472 6,338 7,899 8,738 8,814 8,414 1,976 616 98 306 8,189 7,675 9,063 8,981 7,456 8,579 6,935 8,708 3,711 5,850 4,291 6,044 8,941 8,850 9,384 2,720 8,483 7,496 7,016 7,958 7,233 1,771 322 1,391 473 85 121 1,369 2,710 3,285 670 717 805 1,060 271 7,714 — 7,984 66 144 320 452 1,476 276 2,266 55 — 732 133 631 — 609 — 1,380 — 3,060 — 1,281 — . 5,575 — 3,006 — 4,170 15 2,551 25 987 124 781 — 512 — 7,051 342 825 1,857 551 264 2,240 520 1,176 323 1,996 1 666 1,127 414 253 45 1 2 64 76 66 497 2. 1 1 65 150 391 3 45 3 7 24 3 2 19 15 2 24 208 27 2 348 87 22 140 49 14 26 17 9,653 8,680 826 116 321 5 248 1 45 3 5 72 14 . 1 34 50 31 27 5 5 95 10 27 682 30 240 12 1 196 243 1,931 14 1 8 26 1 1,103 309 18 3 77 1 1 6 67 120 80 1 17 1 1 641 — — — — 1 1 — 1 4 — 4 — — — 8 — — — 65 1 — 1 •6 — 1-4 173 We find, for instance^ the, greatest tendency to sectarian fissiparity amcngst the imbellic Drayidians of the south, and the maximum of eclecticism or indifierence, whichever it should be the more correctly termed, amongst the hardy and oft-invaded tribes of the Panjab. In the former, a doctrinal schism, such as the Lingaiat, for example, led to a completely separated section in every caste that joined in the movement. In the Panjab, the sects, though equally numerous, bear no relation to the life of the com- munity, and the Brahmanic and Musalman sections, whether of the same tribe or district, pay their respects to the same shrine. In the eastern portion of the Central Provinces, again, the lower classes, such as weavers and tanners, have joined in large numbers the Kabirpanth and the Satnami, a couple of sects of the sort described earlier in this chapter, which begin with universal brotherhood and end in becoming a somewhat enlarged caste. The former of the two, though not indigenous to this part of the country, is found -but sparsely represented elsewhere, and the latter also is almost entirely local in its development. In 1881 they were shown in the returns as separate religions, like Sikhism and Jainism, but on this occasion they have not been so distinguished, as it appeared that beyond their acceptance mainly by low castes, they had no better title to such treatment than many other well-known modern sects, such as the Wallabhacharya, Madhwa, Swami-Narayan, Sakta, and so on. In the provincial volume, however, the numbers of their followers are given, from which we can presume that they, are not at all falling off, though the increase may not be so large as that represented by the figures for the adherents of the doctrines of Kabir. As regards sect in the Panjab, where the subject was dealt with at a census for the • first time, the remarks of the Superintendent are well worth perusal, and show clearly the Animistic origin of the most popular beliefs, as distinguished from those inculcated by sacerdotalism. In the adjoining province, too, of the North- West and Oudh, the subject has been taken up, but no report on it is at present available. The develop- ment of Brahmanism in Assam is well described by Mr. Gait in his Provincial Report, and seems in the Brahmaputra valley to present the peculiarity of dispensing largely with the Brahman and substituting the eJ0&cacy of the influence of the Giiru, or ■ teacher. The progress and distribution of the small sects of Brahmo and Arya need no more comment than has already been given to them, but before quitting the subject of the Brahmanic religion a few words are necessary regarding the confusion between its statistics and those of Animistic religions. It has been before remarked Ammism. that the application of the two terms in tracts where the hill population is closely connected with that of the plain was very arbitrary. Por instance, in one district, the whole of certain tribes which were shown as Animistic in 1881, appeared as Brahmanic in 1891, though in the next district the same communities were returned under their former appellation, and in neither had any material change been effected in their creed during the decade. Conversely, the opportunity offered by an alternative title was eagerly seized by the enumerators in Central India as a means of stigmatising as outside the Brahmanic communion, if that term be admissible, all the class of village menials, whose touch is doubtless polluting to the higher castes, but whose religion is but a trifle lower in Animism than that of the society they serve. Thus, in estimating the prevalence of Animism in this part of the country, a deduction of probably nearly one-half should be made, referring to the class of people whose beliefs are elsewhere included under the title of Brahmanism. This reduces Central India from its place as second in the order of prevalence of tribal religions, to fifth, or thereabouts, below the Central Provinces and their States, Assam, and the Bengal States, in all of which the definition has been probably applied as accurately as the state of religious practice will at present allow. In the extreme east and west, that is, in Burma, and in Bombay, Sindh, Baroda, and probably Rajputana, the line between the two is very ill-defined, and the distinction, conse- quently, of little value. In Ajmer and the North-West Provinces the tribes shown as of ' an Animistic form of religion in the neighbouring territory are returned as Brahmanic. In the Panjab, where there is but one tribe thus diversely treated, it is almost certain that the form of religion has really changed in proportion to the distance between the northern settlements of the tribe in question and the hill tract from which it originally came, and where its fellows still reside. The distinction, therefore, is actually in existence, and is not the arbitrary creation of a Brahmanic enumerator. It is not proposed to enter into any discrimination between the different forms assumed by what is here classed under the general head of Animism. There are certain layers of belief throughout the community which are all susceptible of classification into groups now recognised in ethnology, some of the items being tribal, others local, in their prevalence. The task of examination and description has been x o 174 undertaken by various investigators at different times, and Bengal, the Central Provinces, Berar, Bombay, and a portion of Assam have been fortunate in the results, but no general collection of the results has yet been attempted, nor can it be successful until the inquiry has extended over the whole field open to such a survey. Steps in this direction are now being taken by Messrs. Gait and Orooke, in Assam and the North-West Provinces respectively ; and in the Panjab, where Animism in its tribal form has been beaten out through the whole community, so that it can no longer appear under that title, the subject has been well worked by Mr. Ibbetson and others. Musalmans. The next religion to come under review is that of Islam, which is taken here on account of its numerical importance. The Musalman population of the world has been roughly estimated at various amounts from 70 to 90 millions, so that whatever the real figure may be between those limits, the Indian Empire contains a large majority of the followers of the Prophet. No Province or large State, and probably few districts or other subdivisions in the plain country weat of Burma, is without a certain number of Musalman inhabitants. We find them relatively most numerous, of course, in the north-west, where Sindh and Kashmer head the list, with 77 and 70 per cent, respectively. In the former there is a considerable foreign element, consisting of Baldch and Brahui from across the frontier, but the bulk of the population has been converted from a lax form of Brahmanism. Por a short period in its history the province was under a Brahman regime, centred about Haidrabad, where it was disturbed and afterwards confirmed by Alexander the Great, but was overthrown not long afterwards by one of the numerous waves of Skythian origin that broke upon the west and north-west frontier of India before and shortly after the beginning of the Christian Bra. According to the Chinese pilgrim Hiuen Tsang, Sindh was in the seventh century both barbarous and superstitious, and orthodoxy of any sort sat lightly upon its unscrupulous population*, as it is said to do even now. In Kashmer, the present population, whether Skythic or Arya, has been addicted within historic times to serpent worship, Buddhism and Brahmanism, by turns, before its conversion to Islam was undertaken by the Moghals during their summer visits to the valley. The Sikh rule succeeded, but left both Musalman peasant and Brahman professional alike untouched, except that the privileges of the latter were confirmed. • In the north-west the tribes were probably converted from the side of Afghanistan, not from India, and their example was followed by the Mongoloid Thibetan races to the north along part of the upper Indus. On the east, however, in Ladakh, the sparse population is still Buddhist, and along the south range intervening between the valley and the Panjab, there is a considerable Brdhmanic element of comparatively pure Arya descent, but, on the whole, 70^ per cent, of the population of the State is Musalman. In the Panjab we have samples on the largest scale of both foreign immigration and local conversion. In the British portion of the province, 55f of the population professes Islam, the proportion rapidly rising towards the west and gradually falling as the Jamna is approached. On the States, the largest of which, with one exception, are under Sikh rule, Islam has made, of course, less impression, and it is returned by only 30 per cent. As has been said above, the outward , observances of the faith are more or less strictly regarded where the religion is that of a large majority of the people, but left in abeyance where the conversion was effected by force or worldly pressure and without the example of foreign zealots to sustain devotion. The more martial races are converted to the extent of at least one half, and the lowest class of the Brahmanic community favours alternatively Islam and Sikhism. Passing eastwards, we find the proportion of Musalmans high" in the submontane tracts of the North- West Provinces, but below the average in the province as a whole. In Bengal, as we had occasion to note in connection with the density and migration of the population, there is a strong Musalman element, exceeding one half the population, nearly all over the whole of the eastern division, and the ' same remark applies to the Surma Valley, now included in the Assam Province. It is in this part of the country that the results of conversion are more marked in the circumstances of the population than anywhere else in India. We have seen that the gj'owth of the population here has been more rapid than in any other division of the province, and the Provincial Census Superintendent attributes this in a great degree partly to conversion and partly to other ecclesiastical factors as thoy are understood in India. In the first place there is the rise in status, then the range of diet is *" The cattle-breeders are of an unfeeling and hasty temper, given only to bloodshed. They have masters, but shave their heads and adopt the mendicants' robe." 175 greater than amongst tlie Brahmanic classes. Thirdly, not only is marriage deferred till the bride is grown up, but there is no prohibition of widow-marriage, both of which are facts tending towards a longer life on the part of the women and a healthier offspring. In connection with this part of the country, we may mention the Musalman population of Lower Burma, which is largely indebted to Chittagohg and its neighbourhood for its recruits, chiefly sea-faring people, supplemented by a certain influx of the trading Musalmans o^ Bombay and Madras, and the followers of the last Dehli princes, who were assigned a residence at Eangoon. The high proportion of Musalmans in the Bombay States and in Baroda is, in its turn, partly due to the number of traders in Kachh and other G-ujarath , States, partly to that of the cultivators mentioned already — ^both foreign converts — who abound in that division o:^ , the Presidency. It must be remembered, too, that Gujarath was the seat of a considerable Musalman power in the days of Moghal rule in Upper India, and Cambay, Junagadh, Palarpur, Eadhanpdr, and Balasinur, testify to the extent and durability of its authority, as Sachin and Janjira do to the influence of the same religion amongst the foreign employes of the Maratha chiefs. We can now turn to the parts of India where the proportion is the lowest. The Nadir of Islam, numerically speaking, is found in the Hill States under the Central Provinces, and, with the exception of tfpper Burma, where there are now fewer immigrants of the Panth^ class, in those provinces themselves. Next to these comes Mysore, in spite of its former period of Musalman rule. In fact, it nowhere appears that rulers of that faith, when detached from its centre in Upper India, surrounded themselves with large numbers of their co-religionists. Witness the case of Haidrabad, where, though the whole administra- tion is in the hands of Musalmdns, less than 10 per cent, of the population is of that faith, or only 2 per cent, more than in the neighbouring province of Bombay. The proportion in Madras would be very low, as it is in other parts of Southern India, were it not for the trading community of the Labbe on the east coast, and the semi- trading semi-agricultural Mappila of Malabar and its two adjacent States ; for the local convert, in spite of the zeal of Tippoo, is not a considerable feature in the general population, though he is in sufficient force in the larger towns, as shown some years back at Salem, to present a strong front to infringement on what he considers , his privileges by Brahmanical neighbours. The Musalman element in Central India is singularly low, seeing that it was the refuge for many years of wandering bands of marauders of considerable strength. But the strong hand of the two great Maratha powers and the exclusiveness of the Brahmanic chiefs of comparatively pure race m the south-east of the Agency, tend to confine the foreign religion to the Musalman States, of which only one, Bhopal, is of considerable size. As regards the progress of the faith of Islam, little need be added to what has been already AA-ritten above. It has been undoubtedly rapid in Eastern Bengal, and has been perceptible, though on somewhat an ■ uncertain basis, in the Panjab. Elsewhere, the increase seems to be mostly that due to normal growth. But so far as regards the large and heterogeneous class of urban Musalmans found all over the country, it is possible that that growth may have been actually impeded by the difficulty found in getting a living under the new conditions of British ]?ule. For the minimum of literary instruction required now as a passport to even the lower grades of middle-class public employ is decidedly higher than it used to be, whilst the progress of learning amongst this class of Musalmans has not proportionately advanced, and with the comparatively small number of recruits for the army, police, and menial offices that is now found sufficient, few outlets remain available. It is possible that some such reason as this accounts for the fact that the generail rate of increase outside the tracts above mentioned is a little below that found to prevail amongst the population as a whole. In the course of the general sketch of the history of Buddhism, it was mentioned Buddhists, that outside Burma and the Himalayan valleys and table-lands this religion was no longer prevalent in India. The few specimens we find of it in the returns for Madras, the Central Provinces, the Andamans, the North-West Provinces and so on, are chiefly from the jails, where Burmese convicts were drifted temporarily to relieve the local prisons during the days of dacoity. There are also Chinese shoemakers and carpenters, found at Calcutta and other large sea-port towns. Finally, we have from one district in Bombay a relic of the time when in statistical returns Jains and Buddhists were combined, for a small colony of the former were set forth as the latter by somo enumerator whose memory was apparently more accurate than his eyesight, or his perusal of his instructions. The increase of the religion, as a whole, is determined, of course by tbe actual growth of the population of Lower Burma, which is, 3,9 we have y 4 176 •Tains. Sikhs. Zoroastrians or Pdrsis. seen, the most rapid in rate of any part of the country. Ill the Himalayan tracts, too, a little more accuracy in regard to Thibetans has been observed, but still, it is said, d good many of the Bh6tia, or Thibetans of the northern uplands of Kumaon and fararhwdl, return their religion as Brahmanic, when they are in the submontane grazing g-rounds for the winter. The Jains are widespread over India, though they form an appreciable numerical element in the population only in Eajputdna, Ajm6r, and "Western India, and nowhere reach 5 per cent, of the total. It is worth notice that they seem to flourish most where they have devoted themselves to trade and eommeirce, and are weak in numbdt" where they have become agriculturists. In the latter capacity they are found in Madras, but only to the number of 27,400, in Mysore, where there are 13,300,, whilst the Bombay Karnatak returns about 46,000, the bulk of whom are evidently cultivators.' This is a very different state of things from fk^t of the early ceq-turie's of the Christian Bra, when the ancestors of these communities were leaders of lltera.ture and art in the south. Of the trading communities a good deal has been said already in different parts of this work. There are large cplonies ^Ibng the Jamna and in parts of Central India, but except in Dehli, they do iiot much effect the Panjab. In the Gujarath division of Bombay they flourish exceedingly, with headquarters in A.hinedal)ad and settlements at' their mountains of pilgrimage in the Eathiawar peninsula, FalithanS and Girnar. Between them and their co-religionist^ in Upper India come the best known, perhaps, of the whole body, called by the genel'ic name of Marwadi. In Assam and Upper India they are stigmatised , as Kaya,' a term of uncertain etymology, but vulgarly attributed to their inquisitiveness, as it has an , interrogative solind about it. Perchance it will occur to philologists a,t some future peribd" to connect it with the nick-name Kaitoukeitos, given to one of the Deipnosophists of Athenaeus, in consequence of bis aggressive thirst for information. The Jains of the Central Belt and the north Deccan are of this race as well as of Grujarathi origin. It is remarkable that in the country of its birth, Parasnath in South Bihar, there should be no more than 1,487 of this religion returned at the census. From evidence indirectly afforded by applications made from the neighbouring tract just before the census, it seems highly probable that in this part of the country, instead of desiring. to emphasise the distinction of tbeir religion from Brahmanism, as was the case at Dehli, &c., the Jains are anxious toi efface it, as their social position is evidently based on caste orthodoxy within the Brahmanic fold. If this tendency be true, it will account for the disappearance of the Jains into the general sea of Hinduism. The Sikh, religion may be considered as localised in the Panjab, for though there are members of this faith, in most provinces, 98 per cent, of tbem are returned from its , birthplace. The bulk of those in the North-West Provinces are colonists or soldiers, and in the rest of India and Burma, the latter only. In Kashmer, of course, there are some belonging to the Panjab, who have been permanently settled in Jammu or the valley ever since the time of Golab Singh, or of the Sikh domination that preceded the assignment of tlie State to ,him. The Superintendent of the Panjab census shows in his report that in British territory the Sikhs have increased by 24 per cent., whereas in the States under that Government, their numbers have declined to the extent of 19 per cent. He attributes the latter fact to greater accuracy Of definition only, and it is probably due in great measure to the exclusion of the Brahmanic classes returning the Nanakpanth, &c. as their sects. Undoubtedly this is the cause of the notable decrease in the number of Sikhs returned from Sindh, whpre there are few, if any, of the true, or Govindi, Sikhs, outside native regiments. Mr. Maclagan summarises the effect of the change of definition as tending to increase the Sikhs on the return where they are few and to diminish them where they bear a high proportion in the population. .As the Sikhs appertain specially to the Panjab, so the Zoroastrian religion is almost confined to the Western Presidency and States surrounding it. The early settlements of the Parsis atNausari, in the Barpda State, and in. Surat and Broach,. gtill contain about 30-4 per cent, of the entire community, and their original fire temple at Udwada on the Surat coast has maintained its supreme repute. But the head quarters of the race have been gradually shifted to Bombay, where there are now 52-8 per cent, returned. Colonies of some strength are also to be. found in Karachi, Aden, the Haidrabad State, and in P.oona. Elsewhere, small detachments, sometimes consisting of one or twp families only, are settled in military stations, or along the lines of rail. In Upper Burma, the flag was , soon followe,(J by.th© Parsi, ^n^^the / I enterprising Zoroastrian settled himself down iu Mandalay on its occupation with 'vltC same promptitude and confidence in British protection as his forerunners showed in Aden and Quettah. The only province into which he has not yet penetrated seems to be Assam. Here, apparently, trade is not yet sufficiently extended to invite competition, and perhaps, too, the small and isolated European community, for whom the Parsi is always glad to cater, has placed co-operative institutions already in the field. The numerical progress of the Parsi community is a matter of some uncertainty, owing to the migratory habits of the race. Unhampered by caste rule as to food or clothing, the Parsi is free to seek employment all over the world, and many leave Bombay for years together, to work in China, England, or even more distant countries. The census Superintendent for Bombay finds in the age tables reason to think that in 1881 the young children were over counted, so that the increase is really greater than that shown in the tables. On the other hand, temporary causes are manifested in a small community with unwonted prominence, and the conditions of the case, apart from emigration for a certain time from the native country, seem to point to the probability of a slow growth,. The community is a small one, and in spite of its general prosperity and the probable infusion of fresh blood from time to time from various local sources, the marriage field is a restricted one, and domestic ceremonial which is very strictly observed, weighs with undue severity on the weaker sex. Hence, especially amongst the wealthier families, who all belong to the professional and higher commercial classes, there seems to be a tendency towards deterioration in both prolificity and physique, which is not counteracted, as in the case of the middle and the rural classes, by outdoor life and the relaxation of caste rule amongst the latter, or by the energy and success with which the former have of late betaken themselves to athletic sports and exercises. Of the Jewish community, as in the case of the Jains and Sikhs, there is little more jews, to be added to what has been said above. The returns do not discriminate between the various classes of the race, so we are not in a position to exactly demarcate the foreign element from the local. The chief seaport towns, such as Bombay and Calcutta, contain a small and wealthy community which came thither from Baghdad or other places in Turkish Arabia, but which is now domiciled in India. The Aden colony, too, is domiciled there, and seems to flourish and multiply greatly under local auspices, as in 1881 there were but 2,121, and 10 years later the number had risen to 2,826, an increase of 33 per cent. Lastly, there are the two purely Indian com- munities of the Beni-Israel, on the Bombay coast, and the Cochin Jews, further south. The former, in which the local element predominates, shows a tendency to migrate, but does not increase very rapidly. In Malabar, the white Jew, with his small endogamous circle, is apparently stationary, or even on the wane ; nor does he travel far from the places where he first settled, whereas the Beni-Israel is more enterprising, and forms a small, but intelligent and valuable, fraction of the native army in the Western Presidency. In the returns of the Christian population, the distribution by race is given, in Christians, addition to that by territory and denomination, and it will serve as a sort of intro- duction to the rest. From Supplementary Table A, at page 496 of the first volume of the returns, the community will be seen to be composed of 89-1 per cent, of natives of India, including a few negroes ; 7-4 per cent, of Europeans, including Americans and those from Australasia, and 3-5 per cent, of a mixture of these two, known by the general title of Eurasian. The foreign element is composed of, first, the military, next the civil employes Europeaus. of the- State, thirdly, those engaged on railways or mines, and then the professional and commercial classes chiefly congregated at the seaports and provincial capitals. There was also, at the time of the census, a considerable contingent of seafaring people on board vessels in harbour. It is not practicable to separate these groups in the general return. Speaking roughly, the military section amounts to a little over one- half. At the time of the census the strength of the European troops was about 67,800, with 2,550 officers, 3,120 women, 5,900 children, and approximately, 800 people engaged on the staff and in various military avocations in India. To these can be added about 1,200 wives and children of officers and those on the staff, &c., with 2,530 European officers attached to the native army, and about 1,100 persons representing their families, making in all about 85,000 souls. As to the other classes, we have still less of a foundation to build on, A recent estimate gives the civil employes and their families 10,500, and the railway servants 6,100, leaving about 66,400 to be distributed over the other groups, but this is admittedly no more than I 78388. 2 178 Eu Indian converts. Town. European Population. Bombay Calcutta Madras ; Rangoon Karachi - Total 11,458 11,914 4,200 4,120 3,182 34,874 an approximation. Then, again, the distinctions between the three races, is very shaaowy, and there is a tendency for Eurasians to enter the European group, an.d for native (Christians to be returned as Eurasians. As the return stands, however, the influence of the army is visible in the number of Europeans found in the Panjah, which heads the list. On the other hand, Bombay, which follows very closely, there being a difference of less than 200 people, is recruited chiefly from the non-officig.!' civil element, including seafarers. The North-West Provinces and Oudh come third, owing, again, to large garrisons, such as those at Meerut, Lucknow, Bareli, &c. Bengal follows with a fall of nearly 4,700. In the case of this Presidency the civil element is relatively considerably higher than in Bombay, as there are no large garrisons, and, in addition to the mercantile and professional population, and the large port of Calcutta, there is the planting colony in Bihar, Bast Bengal, and the sub- montane tracts, to be included. Madras and Burma come next, though at the very wide interval of 9,800 in the first case, and of 10,900 in the second. In Burma both trading and military population prevail. In Madras the latter is less prominent, perhaps, and the resident European more so. The number of Europeans in places like Quettah, Haidrabad, Mysore, and Central India depends obviously on the strength of the garrison in Beldchis- tan, Sekandrabad, Bangalore, and Mhow respectively. The marginal table gives the European, population of a few of the chief towns, exclusive of the stations purely military. Assuming the Eurasian return to be correct, Madras heads the list with 26,600 out of the total of 79,790. Bengal, which comes next, returns only 15,000. Bombay shows 8,500, amongst whom there arie no doubt a good many Goanese or others with Portuguese patronymics. The North-West Provinces have always been remarkable for the number of their Eurasians, and Burma, which, in spite of its small population has nearly the same number, owes its pre-eminence to its peculiar domestic institutions which foster the supply of children of mixed race. Haidrabad and Bangalore follow Madras in the frequency of the entries of this class. It would be interesting to compare the returns of the two last enumerations, so as to see if this community is numerically on the rise or not, but owing partly to defects in the schedules filled up, partly to the number in 1881 not filled up' at all as regards race, the census operations give little help in this direction, and the matter can only be satisfactorily dealt with through special investigation. Since the community is mostly congregated in the chief towns such an inquiry is not difficult, and can be made to include within its scope valuable details, such as the number of children per family, the age of parents at the time of marriage, and so On, which are impossible at a census. The bulk of the Christian community is, as we have seen above, of native Indian origin, and out of the 2,036,600 souls of which it consists, no less than 1,638,800 are found in Madras and the Malabar States. It includes, also, 152,500 in Bengal, 122,600 in Bombay and Sindh, and 101,300 in Burma. The rest are scattered about in comparatively small societies, such as 23,400 in the North- West Provinces, 19,600 in the Panjab and about 43,400 in Haidrabad, Mysore, and Coorg, where the Madras influence makes itseK felt. Its greatest development is thus found where the Brahmanic caste system is in force in its fullest vigour, in the south and west of the peninsula, and amongst the Hill tribes of Bengal. In such localities it is naturally attractive to a class of the population whose position is hereditarily and permanently degraded by their own religion, as Islam has proved in Eastern Bengal and amongst the lowest class of the inhabitants of the Panjab. We have seen that in the early days of Portuguese missionary enterprise, it was found necessary to continue the breach that Brahmanic custom had placed between certain grades of society and those above them, but in later times, and in foreign missions of the Reformed Church, the tendency has been to absorb all caste distinctions into the general communion of the Christianity of that form. The new faith has thus affected the lower classes more directly than the upper, who have more to lose socially, and less to gain, and a good deal of its early spread in the Malabar States seems to have been due to the success achieved by! the missionaries in their efforts to mitigate the rigour of the local rules and customs 179 towards those classes.* In Burma, again, the chief success has been amongst the Kardn rac68, who are dwellers m the hills and forests, of Animistic proclivities, who, before ithe arrival of the British in this province had been grievously oppressed by the Burmese. The instruction placed within their reach by the missionaries enables them to take their place m society on a level with their former tyrants, who refused them that means of rising ; but though Christianity has, doubtless, got a footing amongst these tribes, the present conditions in Burma seem favourable to the extension amongst the .Karens of the local Buddhism rather than of a more advanced form of belief. It is a question whether amongst the converts in the south-east of the Indian peninsula caste has not still a considerable attraction, for we find that when one of the lower caste converts has prospered in his worldly aflFairs, he manifests a tendency to dub himself one of the warrior-caste of the Brahmanic system, and to appear in the census returns as a Christian Kshatrla ; and in another part of the country the same class, on conversion, considers himself justified in attempting to avail himself of wells, the use of which had been prohibited to him by public opinion whilst he was in his unregeherate stage, thereby causing a village commotion. For the same reason as affects the return of Eurasians, namely, the number of Christians not returned by race in 1881, there can be no estimate formed from the census figures of the general increase in the ranks of native Christians on the last 10 years. Taking special localities, such as the south-east of Madras, the district of Lohardaga in Bengal, and of Sialkot in the Panjab, the variation is very large, and due probably exclusively to the growth of the active community. But even with these remarkable cases of increase ,the Christian native is but in the ratio of seven in a thousand of the whole population, and of these about five would belong to the older churches and the balance to the reformed. The remark, accordingly, made by the Eoman Bishop of Agra to Jacquemont, is as applicable now as it was in 1828 : " La caldaja e molto grande,' ma la eame 6 molto poca." The return of Christian sects is not altogether complete, for we find in the original tables no less than 60,350 persons who omitted to fill in this column of the schedule, and 60,700 more who returned their sect simnly as Protestant. "With the help of the volume of Missionary Statistics before mentioned, these figures were finally reduced to 9,350 and 14,200 respectively, as shown in the following statement : — f Denomination. Total " ^ [ 1 — . Distribution by Race. Number. European. Eurasian. Indian. A.— Church of England, with Churches of India, Ireland, 340,613 103,145 29,922 207,546 America, Anglican, and Episcopalian churches. B._,Church pf Scotland, Presbyterian, United Presbyterian, 46,351 10,581 2,440 33,329 Reformed Presbyterian, American or Irish Presby terians, Irish Presbyterian Mission. C— Baptist a02,746 2,907 2,352 197,487 D. Wesleyan, Wesleyan Methodist, Methodist, Primitive .^2,123 5,362 2,349. 24,412 Methodist, Episcopalian Methodist, Bible Christian. E. — Congregationalist, London Mission, Independent, Calvinist, 48,036 1,285 742 46,009 Welsh Calvinist. ^ IT. — Nonconformist, Dissenter, Puritan i5sr . 230 180 147 G. — Plymouth Brethren, Open Brethren, Swedenborgian, ■ 1,445 178 38 1,229 New Jerusalem, Catholic Apostolic, Quaker, Friend, Salvationist, Anabaptist. H. — Lutheran, German Mission, Swedish Church, Eeformed 69,405 1,285 195 67,925 Dutch, Zwinglian, Moravian, German Church, Evan- ' gelical. Evangelist Church, Evangelical Union, ; Beformed Church. J.^^Protestant 14,213 3,938 4,052 6,223 Total, Reformed Church 755,489 128,912 42,270 584,307 K.— Church of Eome 1,315,263 35,645 36,089 1,243,529 I,. — Syrian Church {Jacobite Section) 200,467 15 3 200,449 M. — Greek, Abyssinian, and Armenian 1,258 726 275 257 Total, Older Churches 1,516,988 .36,386 36,367 1,444,235 • Uqsectarian and Unreturned 11,903 2,702 1,153 8,048 Grand Total 2,284,380 168,000 79,790 2,036,590 Christian denomina- tions. * Linschoten, for instance, writes " As these Nayros (Nairs) go in the street they use to cry ' Po, Po,' '' which is to say ' Take heed, loot: to yourselves,' or ' I come, stand out of the way,' for that the other sort '' of people, called Polyas, that are no iSTayros, may not once touch or trouhle one of them for, if " aiiy of the Polyas should chance to touch their bodies, he may freely thrust him through, and no man ask " him why he did it." There are also strict sumptuary laws regarding the dress and habitation of this class, which are only now in abeyance during the rule of the British, and would be reimposed, with the acquiescence of those subjected to them, within a week from the removal of the foreign hand. See Mr. Mateer's Account of Travancore. — J.A.B. t The grouping was suggested- by Sir T. C. Hope, K.C.S.I., whose help in this portion of the work I have the pleasure of gratefully acknowledging^ , ■■ •■ Z-2 ISO Denomination. Total Christians. By Baccs. ' 3 H c ^ a Church of England Presbyterian IJritish Nonconformists - Lutheran Protestant, unspecified Total, Reformed Roman - Syrian, &c. Total, Old Church Others and unspecified 14-9 61-4 37-5' 2-0 6-3 3-1 124 5-8 7-0 3-0 0-8 0-2 0-6 2-3 5-1 32-9 76-6 52'9 oi-a 21-2 45-2 8-7 0-4 0-3 66-3 21-6 45 '5 0-8 1-8 1-6 10' 1' 13- 3-3 0-3 28-6 61-0 9-8 70-8 0-6 The most striking feature of the return is that 1Q\ per cent, of the Europeans return some denomination of the Reformed Church, whilst 71 per cent, of the Indian converts appertain to the Roman and Syrian denominations of Christianity. It may also be noticed that the last term includes a considerably smaller number than in 1881, the reason being that in the latter year the section of the Eastern Church that had been combined with the Roman on the arrival of the Portuguese was shown under the same head as the Jacobite section, which has remained independent. The marginal table gives the above statement in a proportional form, so as to show the per-centage of each race re- turning each denomination. The Eurasians are fairly equally divided between the two sections of the Church, but the proportion of the unspecified^ such as Protestants, is relatively higher than in the case of the two other races. The ratio of the Roman to the total, in the case of the Europeans, depends first upon the strength of Irish in the regiments serving in India,, and next upon the number of Goanese, &c. who have been re- turned as Portuguese. Next to the Church of England, the native Christian shows a partiality for the Baptist denomination of the Reformed Church, which seems to have achieved its numerically greatest results in the northern districts of Madras and amongst the Karens in Lower Burma. The latter field has been in the hands of this denomination for many years, but the latter received its great impetus during the famine of 1876-78, where it afforded a refuge for orphans and other destitute persons. The Church of England societies are found in most parts of India, but by far its largest congregations are in the southern districts and the Malabar States of the Madras Presidency. There is also a nucleus of this denomination in the hill tracts of Western Bengal, and in the Karen tracts of Burma. The Congregationalists, too, are mostly found in the Malabar States and in North Madras, whilst the Methodists chiefly haunt the North- West Provinces. The Lutherans, who, as seen above, were the first of the Reformed Churches to systematically take up the evangelisatipn of the natives, have nearly 68,000 converts, chiefly in Madras, where they began, and amongst the hill tribes of Bengal. A branch is settled also on the western coast to the north of the Malabar missions. In fact, as a rule, the difi^erent bodies of this section of Christianity seem to map out the field pretty definitely to meet each other's convenience, though there are, no doubt, instances of aggression. We find, for instance, that the newly arrived " Salvation Army," wherever it is in force, is in or near the nests of an older denomination, whose " fort " has been duly stormed like Jericho of old, by the aid of the instrument which, as we have seen, is the most popular one in India, and whose efficacy in the service of sectarianism is attested by no less an authority than Hudibras.* The Roman Church, though the great bulk of its native adherents are found in Madras and in the neighbourhood of that Presidency, has a strong foothold in Burma, but its greatest success during the last decade has been in the Lohardaga district of Bengal, where the wild tribes came over in numbers during an agrarian crisis a few years ago. In other parts of Bengal, too, as well as here, the census Superintendent points out that there have been reprisals from the Baptist and Lutheran denominations to a considerable extent. The Groanese element, together with the curiously isolated community of native Christians on the island of Salsette, near Bombay, is the main factor in the Roman Church of the Western Presidency. It is worth while to compare the returns furnished to the census enumerator with those in the missionary publication already more than once quoted, though this can only be done in a very general way. The figures for the Roman Church agree very * " We'll beat a drum, and they'll all follow. A drum, quoth Phoebus, 'troth that's true, A prptty invpntion, quaint aiid new I " 181 fairly with those given in the year book called " Missiones eatholiGEe," the ditferenoe being a trifle in excess on the side of the census. .In the case of the JReformed Church, too, the total results are biassed in the same direction, but in some of the details there appears a great difference. In the course of the preparatory operations before the enumeration, a circular was issued by the Superintendents to all missionary agents suggesting that the Societies concerned should issue cards to all native householders of their denomination, with the name of the latter legibly entered thereon in the vernacular character of the district, so that it might be copied into the schedule with as little danger of error as possible. This course was evidently not universally followed, and in default of such a guide both enumerator and enumerated were at a loss to get at the correct entry. The neglect of such a precaution seems to indicate a dislike on the part of the leaders of the congregation, similar to that which prevails in Great Brita,in, to the return of denomination. On the other side of the question it ma.y be pointed out that a good deal of the missionary statistics is admittedly approximate only, and in some cases is based on the simple rule of multiplying the number of communicants by 2^ to get that of the whole denominational community. Taking Indian converts only, we find the main differences between the two returns to be as shown in the mar- ginal table; some of the discrepancy is attributable, no doubt, to defective no- menclature. For instance,, the group marked F. should be distributed, apparently, in the case of the mission return, between Lutheran and Baptist, and the census return under group 0. includes, probably, some Congregationalists. The Salvation Army, again, claims considerably more than are assigned to it in the census schedules, and may possibly be confused with the Presbyterians, alongside of or amongst whom it works in some parts of the Bombay Presidency. As regards the variation from the 1881 return, the scrutiny of denomination shows that the only figures coming under identical heads are those for the older churches, which, too, omitting the small number of the Grreek and Armenian sections, must be taken en masse, on account of the confusion between the Ebman and the Jacobite sections of the Syrian Church, which occurred in 1881. Thus taken, the increase is about 19-2 per cent. If the above items be deducted from the grand total, the increase of the remaining denominations, representing the Reformed Churches, amounts to 28-6 per cent. The " Protestant" and other unspecified entries numbered 189,692 persons in 1881, and 142,925 in 1891, a difference which affects a considerable proportion of the detailed entries of denomination. It should also be menfcioned that there is a discrepancy between the number of Christians by race, as shown here and as shown in Table XVII. In the case of Europeans, the difference amounts to 1,572 in favour of the return under review, whilst the Eurasians are here shown as fewer by 1,954. The discrepancy is chiefly in Madras and Rangoon, and is apparently attributable to the revision of the tables after the Imperial series had been completed. The entry of Goanese as Portuguese, and the return by Eurasians of non-Christian religions accounts for part of the confusion. As to the Indian converts, w^ho are here returned in excess of the other table by no less than 200,742, the explanation obviously is that in the latter there is no distinction by religion shown, so that much of the Christian community has been compiled under the caste or tribal name which they returned in the portion of the schedule assigned for that detail. We have mentioned above the case of the toddy-drawers of the south of Madras, who return themselves as Christians of the warrior caste of Brahmanic orthodoxy, and many other Christians will be found under the lower castes. These are mentioned in the provincial reports, but in accordance with the general scheme of the Imperial return, the distinction is ignored in the latter. Moreover, ^s there was no requisition on Z S Denamiuation. A. — Census. B. — Mission. C— B. com- pared with A. A. — Church of England 207,546 194,482 -13,064 B. — Presbyterian 33,329 34,979 + 1,650 C— Baptist 197,487 133,762 -63,725 D. — Wesleyan, &c. 24,412 33,058 -1- 8,646 E. — Congregationalist 46.009 78,135 + 32,123 F. — Indefinite Nonconformist 147 23,794 +23,647 G. — Miscellaneous {chiefiy Sal- vation Army). H. — Lutheran 1,229 67,92.5 2,491 63,351 + 1,262 - 4,674 J. — Protestant 6,223 — — 6,223 Total 584,307 563,949 -20,358 182 Christian converts to return the caste to which they belonged before their conversion, and the general feeling is adverse to the recognition of thiat caste, except in special instances, chiefly found on the Malabar coast and near Bombay, the return is necessarily only partial, and most of the community appear under the collective title of Christian. Nature aud origin of caste. C. — The Distribution of thk Population by Race, Tribe, or OaSte. YlaTDioog TTOLpiho^ag dg Q^ o[xri\tKOig XP°^*^ Ke/CTV]j«.e6' nu^eig dvrt Kara.fia.'kei "Ktyog, 'OoS' e< Sj' OLKpov TO o-oc^ov TJupr^rai ^p6vc7M — Euripides^ In this section we have to. deal with the last , and the most important of the ethnographic elements included in the census scheme ; and, it is also, unfortunately, that of all others which it is most difficult to treat adequately within the limits of the present review. Philology, as we have seen above, is of little help in the demarcation of. the ethnic distinctions amongst the people of India. As to religion, again, what has been said already is enough to show its imperfection as a guide, and of all the erroneous, methods of discriminating between the divisions of the community, that which assigns any ethnic value to the terms Hindu and Musalman is the most mis- leading. It is true that very high authorities have described caste as the " express badge of Hinduism," and M. Barth, whose testimony has been already cited more than onc^ in the course of this chapter, considers that this institution is not merely the symbol of Hinduism, but its stronghold, and a religious factor of the very highest order. To a very great extent, so it undoubtedly is, but what has been said in the foregoing pages justifies, perhaps, a wider view of caste, by which term, it is as well to explain, is here meant the perpetuation of status or function, by inheritance and endogamy. For caste is a development of the special tendency to which the social atmo- sphere of India is abnormally favourable, and is not, therefore, the peculiar attribute of the Brahmanic form of religion, whatever it may be, so much as of the circumstances of which the Brahman had the opportunity of moulding in days long gone by. Por example, we find that where tribal pride is strong, and the social position good, the change of faith involves no alteration in either nomenclature or social custom. Where, on the other hand, the community in question is in a depressed condition in the Brahmanic system, conversion to another faith is used as an opportunity to slip out of the hereditary yoke. Thus we fiud the various tribes of Jat and Rajput, in the north of India, contain nearly an equal number of Brahmanic and MusalmAn members, not to mention the Sikhs that prevail in certain localities. Amongst the Jains, again, of Western India intermarriage is not uncommon between the Brahmanic and Shrawak sections of the same community. In another direction, fiction assigns to the leaders of a forest tribe the position and honours of the warrior caste when they enter the fold of orthodox Brahmanism, whilst 'the rest of the tribe retains its original designation, and sometimes, as we have seen in the preceding section, remains even professedly Animistic, or, as in Assam takes a special title conventionally allowed to those who join the Brahmanic community. The case of the eastern Bengal Musalmans has already, too, been quoted as an example of social rise on conversion, since they all assume the honorific designation of Shaikh, implying admission into the Arab fraternity. The Christian converts of Southern India may be cited since a further instance of this tendency. The Shanan, or toddy-drawers, with the corresponding class on the western coast, hold but a very low position in the Brahmanic system, as it is in force in Dravidian lands. When, therefore, they become Christians, most of them abandon their caste name, but as they rise in worldly circumstances, the tendency comes out, as has been mentioned at the end of the last section, to arrogate to them- selves the title of the warrior caste, like the wild tribes of the Central Belt. But under the early Portuguese missionaries, higher up on the Malabar coast, the distinctions of Brahmanic caste were always recognised on conversion, as the object was to provide the church with a class of convert who was fit by position, rank, and education, to take up an influential position in the priesthood.* ' Numbers of similar instances can be cited of the independence of the obligations of caste from the elastic bondage of what, with the masses in India, passes for religion, a non religando. \.s thit; is going through the press, I see that this point has been noticed in a leading Indian journal in the ng fitting phrase: — "The multiplicity of Hindu castes is such ihat some have even assimilated the * As following fitting phi . _- terminology of the Christian faith."- -J.A.B. 183 But granting that there is something inherent, as it were, in the conditions of hfe in, India that fosters the sentiment of which the caste system is the expression, and grantmg again that the form, or collection of forms, Ihat this expression has taken is the outgrowth of the Brahmanic creed, it seems within the bounds of reasonable I hypothesis to attribute , to its present development an origin distinctly racial. -' We have seen the origm of that creed in the necessity of keeping the gods on the side of the Arya, in the struggle of the latter with the dark races they found m possession of the tracts they coveted. We find the germs of, the caste system m the subsequent contact of the foreign race with their late opponents, when aot^al hostilities had closed. The Brahman had already developed into a hierarphv and by their influence over the supernatural, could easily prevail on the Arya laifcy to exclude the subject races from participation in what seemed to be the peculiar privilege of the superior. There is thus an ethnologic justification for the use of the Vedie term " colour," m the sense in which caste or race is now used, as has been pomljed out by modern writers. But here the racial element seems to cease for the moment, for ipside^ the community of the Arya themselves, assuming that the fulmmations of the Brahman had any effect, the development of function must have proceeded on caste lines. That is, the right to sacrifice, and to perform the other offices of the priesthood having been established on a hereditary basis, the natural tendency would be to extend the principle to those who supported the priesthood, namely the warrior chiefs. So, too, within the circle of the Yaishya, or colonists of Arya descent, function became hereditary, through trade guilds, in imitation of the higher classes, and with the growth of the settlements of the race, the hierarchy separated from the commonalty, so that practice was less supervised, and the non-Arya undoubtedly entered by the lower gate., even though interbreeding was not prevalent till a later age, which is a doubtful assumption. The literature of the Brahmanical revival, after the downfall of Buddhism, is practically devoted to caste, that is, to the support of the Brahmanic pretensions, and on this is based the system that prevails in the present day. The process of absorption of the Hill tribes, mentioned in the preceding section, necessarily tended to obliterate the line of demarcation between the Arya E^shatria and the black chieftains now admitted to the title. Indeed, according to the orthodox doctrine, the whole of the Kshatria caste, which apparently was found inconvenient in the new order of things, was destroyed to a man, by the Brahnianical creation named Parasu-E.ama, but this is a difficulty that no hierarchy would fail to solve.* Then with regard to the Brahmans themselves, who may be assumed to have done what they could to defend the sacerdotal body from recruitment from outside, we read of instance upon instance of the creation of the class by royal edict, and of others where the Brahman, having descended to temple ministration, is degraded from the hierarchy. In one province alone there are over 150 divisions of Prdhmans, none of whom will intermarry with each other or sit down together at a ceremonial banquet. If this be the course of events at the root of the tree, it is hardly necessary to despant upon the chances qf gepealogical purity amongst the branches. The Arya entered India, no doubt, white in colour, and more hardy and advanced in the arts of war and husbandry than his antagonists, but not so far removed in race as to prevent successful interbreeding. Then, too, being an Asiatic, albeit of the temperate zone, he was able to settle in Upper India, beyond the tropic, . with as little inconvenience as regards climaite as the Burman in the neighbouring peninsula. There was no question of physical deterioration, as in the case of the European and those crossed with him in India and corresponding climates. As the settlers receded further from the north-west of the country, their touch of the fathers of their race, and the chances of recruitment from across the mountains, grew less, until there is no basis for even a reasonable pretence of Arya blood. The case is often cited of the importation of a few families of Brahmans into Bengal from the centre of Brahmanic orthodoxy, to restore the decaying practice pf the faithful in the Deltaic marshes, and the consequent repeopling of that tract, even unto the present day, with the pure blood of Kanaujia. But from the original tradition, it appears that only the first wives of these prolific teachers were of the BrAhman caste, and that the rest were daughters of Heth. It is the same in JSTipal, the most orthodox tract in India, and the only one wherein the Musalman has never set his conquering foot. The ruling family was found by the refugee Brahmans to be * The new race of chiefs, according to the Puranic and Epic literature, resembles the Levitic conception, given in the Chronicles, of the Kings of Israel, who war for, and at the command of, the priesthood, with minds filled, says Eenan, with nothing but priests, Levites, and poets. — J.A.B. Z 4 184 of Rajput blood, and the Brahmans themselves laid the foundation of the -well- known Khas tribe, of mixed blood, who now constitute the predominant community in the whole of the lower valleys. In this tract, by the way, it is said that there is the only opportunity to be found in" the present day of judging of Brahmanism where its development has been absolutely unchecked ; and something very like Physiology the code of Manu is in force in its unmitigated bigotry. It is only within and caste. comparatively recent times that the racial distribution of the population has been investigated on the line of modern science. Previous to that time philology had held the field unchecked by observations from other standpoints. A beginning has been made in Bengal by Mr. Eisley, who has published anthropometrical data from about 6,000 persons. Most of them are from Bengal, but some were made in the North- West Provinces and the Panjab. Such an extensive field of survey, com- prising over 146 millions of inhabitants, cannot, of course, be appreciated from the results of measuring one person in 24,000, but the results show the value of the method, and it is to be hoped that more material of the same sort may be made available. It is imprudent to base any general conclusions on this small foundation, but so far as the local circumstances of Bengal are concerned, it may be inferred that the population consists of a mixture of two breeds, first, the western, in which the Arya is an element, especially in Bihar, but mixed with the dark races of the Hill tracts. Secondly, the eastern, where the predominant factor is the Mongoloid of the north-east, mixed with the dark races of the plains and marshes, and with a slight tincture of Aryan blood in a portion of the upper class. One valuable feature in the measurements is the apparent identity they establish between the two families of the dark inhabitants of the Central Belt of hills. The distinction, as a purely linguistic one, is still untouched, but physiology has pronounced against any difference between the Kdl and the Oraon or Grdnd. Taking the measurements of the Panjab for what they are worth, it appears that the lower castes in that province possess a higher type of feature than the higher castes of the Lower Granges valley. This conclusion jumps with what is known of the history of the population of the north-west corner of India, where not only has recruitment from the hardy races of Central Asia been going on within a far later period than any that can be assigned to an Arya movement from the same direction, but, moreover, the dark or Dravidian race does not seem to have ever been prominent in the land of the Five Elvers, so that the development of caste in the latter region, wherever it obtains, must have been mainly tribal or functional. The alternative hypothesis, to which the existence of the village helot in the centre and east of the Province gives colour, is that the race that preceded the Skythic tribes who were found in possession by the Arya were of a northern stock ; or, again, that the race had been long enough in India to change the type in the different physical conditions of the peninsula. But it seems very doubtful whether such migration would affect more than the skull-shape, and would not leave unaltered the otBer facial measurem^ts and the colour. All that can be said safely is that whatever the mixture, the race type of the north and west of India differs from that in the south and east, and that of the Arya type, whatever remains, must be sought in the Ganges Doab. On the west coast there are a few curious distinctions that indicate, apparently, difference in racial origin. The first of these instances is that of the Nair, the military caste of Malabar. Their traditions point to the north as their native land; they are light in colour, in very great contrast to the rest of the castes of the tract, have retained the custom of polyandry, with a good deal of serpent worship. It appears that they advanced upon their present tract by way of the coast higher up, but how they got there does not appear. As with the Arya, they found a dark race in pos- session and enslaved them on their estates, where they labour to the present day. In the same tract, too, there is a class of Brahmans, the Nambudiri, of remarkable fairness of complexion, and noted for their rigid ceremonial puritanism. Then, again, in the track of the Nair's alleged progress, we find a peculiar caste of Brdhmans, partly occupied in the cultivation of spices and betel nut, but settled mostly above the Ghats, and not therefore so well sheltered from foreign influences as the Nair, who sought the coast. These Havig or Haiga Brahmans show their connection with the Ttilu country in their speech, and, like the Nairs, attribute to their caste a serpent origin in Eohilkhand, a statement borne out by their title. • Between these we have a class of female temple servants of an equally light complexion amidst a universally dark population. Further north come the well known Konkanasth Brdhman, sometimes called Chitpawan, who gave to the Marathas their line of Peshwas. Their traditions point to an origin beyond the sea, and they have certainly preserved a complexion and features in marked contrast to those of their neighbours. It is not proved, however, 185 that the sea they crossed, was wider than that betweeu Sindh or Kathiawar and their pi-esent settlements in Eatnagiri, so that they may well be the remnants of one of the really Ary a clans of Northern India; either that harried by Alexander on his voyage down tlie Indus, or that which was the appanage of the Rajput courts of Saurashtra or south-western Rajputana. At all events, these curious instances of an evidently northern community embedded in a purely Dra vidian entourage show that a good many links in the ethnological history of the country are wanting. So far as race is concerned, then, India has a north-western or Skythic element in the Panjab, fringed by the semi- Arab blood of the Afghan and Baldch along the west. It is impossible to say where the former influence ends towards the south and east, but it probably reached as far as the peninsula of Kathiawar, in one direction, and nearly to the Jamna in another, but all along the border line the demarcation has long been obliterated. There is then the Mongoloidic element in the Himalaya and the eastern valleys, stretching down to the sea at the extreme east, and much mixed up with the population of other races throughout the trans-Grangetic tracts of the Delta. Burma may be said to be entirely Mongoloidic. Along the west coast there are traces here and there of Skythic blood, and even of that of races still further removed from India, who have settled there, however, within historic times. In Kashmer, the lower valleys of the Panjab and Garhwal Himalaya, the central portion of the Upper Granges basin, and, probably, the eastern part of Rajputana, the prevailing racial element is as nearly Arya as in any part of India, the tinge getting darker as the hill country is approached. In the latter we have the Dravidian, or whatever the appropriate title may be, and with the exception of the few communities along the west coast already mentioned, this race predominates throughout the peninsula, the hills everywhere standing out with special purity of blackness. It is not possible to give accurate statistics regarding the relative prevalence of these races, owing to the assumption of membership of well-known social orders to which the newly- enrolled have no racial title, nor can any approximation be made that is not likely to turn out fallacious, owing to the great confusion of blood in every part of the country, except Burma and the Hill tracts. The evolution of caste from the comparatively simple basis of race and function DevRlopment into its pi-esent complicated shape is a subject of which it is impossible to give here of caste, more than a bare sketch. The supersession of racial by tribal feeling is not an unusual feature in any country where the two are found, and in India, no doubt, it has been helped by change of religion in the more frequent cases of widespread conversion. Of castes of political origin instances have been already given, and more can be obtained on reference to any of the census reports of the last two enumerations. The most striking example, however, seems to be that of the Hill tracts of the Panjab, where the Raja is the fountain of honour to an extent unprecedented, probably, in other parts of India, for by his word he creates, enlarges, and restricts the castes of the population of his realm. Elsewhere, the process of creation is, as a rule, confined to the two upper classes, Brahman and Rajput. "We have, then, the incor- poration by fiction, as it has been called, of members of the forest tribes, of which enough has been said above. The gradual acquisition of a higher caste by the lower seems not confined to any particular locality or system, for in Assam Mr. Gait has well described it as becoming almost a rule amongst certain classes. In the Deccan, arain the acquisition of wealth by a landholder often carries with it a rise in caste, of which the first manifestations are the seclusion of the female members of his family and the abjuration of widow marriage. In the next generation the family becomes Maratha and then proceeds to Rdjpiit. The Christian aspirations after the latter title have been mentioned already, and in one instance a Madras apothecary of this faith applied for a wall round the yard of the dispensary of which he was m official charge, in order that the women of his fainily might take exercise m the seclusion due to their caste But the great majority of the new formations on social considerations seem to be in the direction of degration rather than of usurpation. Some violation of caste ceremonial or etiquette is met with excommunication by those who have not yet tnpped, and the peccant community at once forms a subdivision of the old, intermarrying 'onlv within itself. In this way, the semi-agricultural grower of fruit, generally a suburban resident, has declined communion with his fellow who devotes himself to the provision of vegetables. The carpenter, again, will have no companionship with him who repairs the carts used in municipal conservancy. Other instances are found scattered throughout the caste history of every province. We then come to territorial divisions, which are often based oh difference of language, The functional pastes, of , 78388- A a 186 course, furnish, tlie bulk of these, and it ,is rare that there is conniiMum or confarreatio between the two sections. Sectarian differences, too, are fruitful-sources of new caste growths. Some instances have been given in the section of this chapter which treats of the development of religion, and the process is pretty generally: on the same lilies in all. There is a sort of revolt against the Brahmanic authority, which, attracts people of different castes. If it involves nothing but dogma or academic principles, caste is not affected. If, however, social ortl}.odoxy is impugned, the adherents ' have to retire from the fold and form a new community. Thus, the Lingaiat Sect of the South Deccan includes a number of different castes, all receivillg separate recognition within the sect, and all prohibited from intercourse by marriage at all events, with the castes from which they have respectively seceded. In a number of cases,' too, the name of the sect was returned as a caste without further discrimination, and, as in other instances, such as those of the goldsmiths and silk-weavers, thb spiritual ;guides of the community have laid claim to the title of ■ Brahman. The communities in question are all in good circumstances, and it is noticeable that whereas 'all the great !N^onconformist movements of India, involving, that is, large numbers, h.ave been in a direction extra Brahmanic, and patronised by the lowest sections of the population, the- numerically smaller have been taken up by the well-to-do' and directed, ultimately, to the advance of the community within the system to which they originally belong* For, to quote Butler again, " Gain has wonderful effects, To improve the factory of sects." Thus, in spite of the theoretically immutable ba,rriers of caste, it is plain that the community which, acknowledges them is constantly leaping oyer or creeping under them, or adding to their number, but in all this change .steadily adhering to the main and vital principles of the institution. It is of the highest interest to trace the various influences that seem to pervade the caste system of the different parts of the country, but space does not allow of it,* and in addition to the chapters on the subject in the provincial and State reports on the census, both on the present occasion and in 1881, works are available for reference, such as LyaiU's " Asiatic Studies," Eisley's work on Bengal castes, Crook's and JNesfield's on those of the North- West Provinces, and' the " Bombay Gazetteer," with many others, from which the chief features can be learned, with much valuable folk-lore into the bargain. We have now to consider the results of the enumeration of caste, and it is necessary, accordingly, to show, as in the case of occupation, what was the scope of the inquiry. The following statement, therefore, gives the heading of th.e schedule and the standard instructions : — 1881. re. Religion. Column J T,,,]- ;„ _J Headings ] «''"g'o°'^ 7. Caste, if Hindu ; Sect, if of L L other Religion.* Column J Column Headings. ^ 1891. Tj„i;„-„„ / 2. Main Religion. Religion (3. Sect of Religion. Caste or / 4. Main Caste, Tribe, &c. , Race. \ 5. Sub-divisioa of Caste, Tribe, &o. Instructions.— Col. 6. Here enter the main reli- gious denomination to which each person belongs, as [Hindu, Muhammedan, Christian, Sikh, Jain, Brahmos, or other religious division of Hindus cot mentioned above, Buddhist, Jew, Parsi]. If a person belongs to some aboriginal or non-Hindu tribe, enter the name of his tribe, as Garo, Khasia, &c. Col. 7. In the case of Hindus, here enter the caste, as Brahman, Kajput, &c. If the name of the sub- division of the caste be entered, you must also enter the general name of the caste, as a whole. In the * In this colnmn such 'details of tribeii, clans, and sects, as the Local Government may direct, should be introduced. Instructions. — Col. 2. Here enter the main religion which each person returns : — >- As Hindu, Musalmdn, Jain, Pdrsi, Christian. Forest tribes who are not Hindu, Musalmdn, &c., should have the name of their tribe entered in this column, iis Bhil, Qpnd, Garo, &c. ; low castes, such as Chamar, Ddm, Pdria, Mahdr, Sec, should be entered by the religion which they themselves return, and no dispute about it is to be raised. Col. 3. Enter the sect of I'eligion followed by each person as they return it : — As Smdrth, Vaishnav, Walabhacharya, Lingaiat, &c^ for Hindus ; Sunni, Shiah, &c., for Musalman ; and for Christian enter whether Church ofEngland,Roman Catholic,Presbyterian, Baptist, Lutheran, Methodist, Episcopalian, &c. If the sect cannot be stated, enter " Not returned '•' in this column, but do not leave it blank. ' * For example, the curious division in the Tamil country of the castes, other than Brahmans and Kshatria into those of the right hand and those of the left, unknown in any other part of India ; and the still more curious custom of the males of the leather workers and females of the Polya (jnenials) beiOff pn the left whilst the other sex of those castes is dignified by a place on the right. During factional warfare' conjiigal' rights are suspended, and resumed when peace again reigns. — J. A. B. v^ *, /s 187 1881. (Continued.) 1891. (Continued.) case of Musalmans, state whether *hey are Sbiahs, Sunnis, Parazis, or Wahibis. Christians should be showu as members of the Church of England, Eoman Catholics, Presbyterians, Baptists, Wesleyans, Ar- menians, or as belonging to the Grreek or Syrian rite, or, if not belonging to any of these denominations, under the general head of " others." Col. 4.* Enter the caste if Hindus or Jains, and the tribes of those who have not castes, and the races of Christians, Buddhists, &c. : — As Brahman, Rajput, Bania, Kukbi, for Hindus ; Pathdn, Moghal, &e., for Musalman ; Eurasian or Native Christian for Christians ; do not enter vague terms, such as Hindustani, Marwadi, Panj&bi, &e. Col. 5.* If the caste has been entered in col. 4, enter here the sub-division, as Kanaujid, Ndgar, &c., for Brahmans, Osvdl for Banias, &c. Jf tribe, enter the clan, if race, enter the tribe or nationality. Some races or castes may not return sub-divisions, and in their case the entry in col. 4 should be repeated, but this column must not be left blank. Native Christians, for in- stance, may be returned as Portuguese, East Indians, Madrasi, or by their caste, if recognised. Karens, as Sgau, &c. ; Bhils as Tadwi, Pdwada, &c. ; G6nds as Raj, &c. * The drafting of illustrations in the instructions for cols. 4 and 5 was left to the local census Superintendent. As to the main points of difference between the two sets of rules, it will be seen that in tiie later code the caste was separated entirely from the religion, and the sect, accordingly, kept equally distinct. In other respects, the difference is merely in the greater detail of the illustrations, a matter that was left largely to the discretion of the local census Superintendent, The first mentioned change was made in order to get rid of the notion that the caste, or social distinction, was not required for Musalmans, Sikhs, Jains, and so on, or was held to be subordinate in any way to sect of religion. With respect to the greater detail, it was thought that with plenty of discretion allowed to those who knew the special ways in which the enumerators were most likely to err, delegation was, on the whole, better than reservation, as in some cases on the last occasion prohibition of certain entries had been found to be taken as prescription. On the whole, too, the separation of the four items of religion, sect, caste, and sub-caste, &c., was found to make the notion of what waS required more easily intelligible to the enumera;tor. Mistakes in abundance there were, of course, and the chronicle of these can be read in the provincial reports, where they are duly noted; but, speaking generally, the detail was more correct than on the former occasion, and ju.stified the greater elaboration of the instructions. Then came the process of abstraction, in which the main chance of error arose from the misreading of badly- written titles, more especially in the case of names of cases foreign to the district, of which alone the clerk had experience. When the heavy task of checking and tabulating the district totals was completed, there remained that of classificationT It will be as well to explain this process by reproducing the Census Commissioner's Circular Order on the subject, as was done with reference to the corresponding process in the case of occupation. On the Abstraction and Tabulation of Castes, Tribes, SfC. The Abstraction should include every caste, &c., and every subdivision, found in the schedule. From the Abstraction sheets an Index is to be prepared in which each caste, &c., with its subdivisions, should aoBcar in alphabetical order. In a province such as the Panjab, where the question of caste distribution was fully dealt with in connection with the last census, it has been found unnecessary to tabulate more than certain locally important subdivisions, after abstracting and indexmg the whole Elsewhere, it will be advisable for a comolete table, showing the strength of each item, to be recorded for further inquiry. foi %<=«™Pf^y^'^;"f' ^^^ jj^^ ^^ been already given in Appendix B. of Circular M and the present Note suDPrements the above instructions with the scheme of classification therein nientioned Generally speaking, tL meThod of grouping is a rather more comprehensive form of that adopted by Mr Kitts m his Compendium of cTstes but, Instead of following clbsely the order and detail of the Occupation able, an attempt has been made to arrange the groups more or less in accordance with the pos.uon generally assigned to each in the s^cid scde as suggested by Mr. Ibbetson in feis Panjab work. A class has oeen added too, for those who do Lot belong to the Hindu, Jain, or Musalman communities or to Forest tribes, which addition will include, «l,o convert! to Christianity who do not retain tiieir original caste or tribe. No sample list has been prepared as there is danger' of confusion between the nomenclature of Different provinces; but Mr. Kitts' nit aSs a general iAdication of the application of the scheme, thpugh no doubt, the detailed knowledge of .L^n^«l Superintendents will enable them! to make corrections in certain points of detail, i he main object Fmo Sure uniformity of classification, so that the circumstances of the provmces m respect to sex-distri- 188 bulJoii, education, and marriage customs, where abstracted, aud tiie prevaleuce of the selected infirmities, &c., may be accurately compared, in spite of the local name borne by the same caste in differtint parts of the continent. 3. For Imperial purposes, therefore, the numerical table required should include every distinct caste or tribe under its appropriate class and group. In several provinces the main castes are divided by religion, so that the latter fact should be recognised, as well as the social distinctSon.* Kajpiits and Jats, for instance, in the north, are found to be, Hindus, Musalmans, and Sikhs. Baniyas are both Hindu and Jain. Forest tribes return themselves as Hindus, and of the tribal religion, and in several important causes as Musalroan and Christian also. Tlie provincial form of this table should, moreover, mention such sub-castes, &c., as may be recorded in tabulation ; and on the subject of arranging and defining these subdivisions useful hints will be found in a paper by Mr. H. H. Eisley, printed at pages 343-352 of Volume I., No. 6, of the Journal of the Anthropological Society of Bombay. In the Imperial Return for India the number alone of such subdivisions is sufficient, and a further selection will be made in the Central Office of the castes in each class or grouj) which will have to be separately named. * This refers to the provincial series of tables, as it may he worth while in special eases to see if conversion works any change as regards education or marriage. — J. A. B. The scheme of classification mentioned in the above circular is as follows ; the serial numbering, however, is that finally adopted for the Imperial Return for India, which differs a little from that originally prescribed, and which will be found in use in the provincial volumes : — I. Military and Dominant. XXXI. Potters. i-s IX. Other Agriculturists. t-3 XXXII. Glass and Lac Workers. 93- III. Cattle Graziers. s XXXIII. Salt and Lime Workers. ' < IV. Meld Labourers, U3 XXXIV. Goldsmiths' Refuse Cleaners. < V. Forest Tribes. 1 XXXV. Gold Washers and Iron Smelters. VI. Priests, &e. i XXXVI. Fishermen, &c. VII. Devotees and Ascetics. > XXXVII. Domestic Service, &c. VIII. Temple Servants. XXXVIII. Distillers and Toddy Drawers. 1 IX. Genealogists. XXXIX. Butchers. o X. Writers. < XL. Leather Workers. ^ XI. Astrologers and Herbalists. 6. XLI. Village Watchmeu, &c. P3 XII. Musicians and Ballad Reciters. XLII. Scavengers. XIII. Singers and Dancers. XLIII. Grindstone Makers. XIV. Actors and Mimes. XLIV. Earth Workers, Stone Quarriers, 1 a 3 XV. Traders. 1 XLV. Knife Grinders. \ XVI. Pedlars. 1- XL VI. Mat and Cane Workers. d 9 ^ XVII. Carriers by Pack Animals. XL VII. Hunters and Fowlers. ' XVIII. Goldsmiths. XLVIII. Miscellaneous Vagrants. XIX. Barbers. XLIX. Jugglers, Acrobats, &c. XX. Blacksmiths. L. Musalm^n Foreign Races. '3 XXI. Carpenters, Masons, &c. lil. Himalayan Mongoloids. 1 XXII. Brass and Copper Smiths. 1 LII. Burmeee and Chinese Mongoloids. XXIII. Tailors. LIII. Western Asiatics. > XXIV. Grain parchers and Confectioners. '3 03 LIV. Burmese, &o. Mixed Races. a XXV. Perfumers, Betel-leaf, &c. Sellers. A- r— j LV. Indefinite Indian Castes. u XXVI. Weavers and Dyers. tn LVI. Europeans, &c. < XXVII. Washermen. 1 LVII. Eurasians. XXVIII. Cotton Cleaners. P"^ LVIII. Native Christians. XXIX. Shepherds and Wool Weavers. LIX. Goanese aud Portuguese. XXX. Oil Pressers. LX. Africans. This grouping, it will be seen, is based mainly upon function, but it was further explained to the classifying officers that the occupation to -which the caste in question was to be credited was not necessarily that actually exercised by the caste in the ^189 present day but that which was assigned to it by tradition, and generally implied in its current appellation. There is thus no connection between this return and that of occupation. ; Again, towards the end of the classification, a set of groups will be found in which race is the predominant feature. It has been explained above that the Musalman titles that appear in Group L. do not, except in the Panjab, necessarily imply foreign origin. In Group LVIII.,, too, the term Native Christian" only covers those of this community who returned themselves under no other title. As pointed out in the circular, what was aimed at in prescribing the classification, was as much uniformity as the nature of the statistics will allow, so that the returns of each province migh't be dealt with on the same basis. This was especially necessitated by the tabulation of caste or race, as the case may be, in combination with such details as literacy, infirmities, and, where undertaken by the local authorities, with marriage and widow- hood also. The index sent up from most of the provincial census offices enabled this object to be fairly fulfilled, but owing to the absence of such a guide in the case of the two largest provinces, comprising more than half the population of British territory, the returns for which did not reach the Imperial office until six months after all the rest had been compiled and worked up, inconsistency could not every- where be avoided. In the summary statement, called Table XVII. — ^A., with which Yolume II. begins, the more important adjustments that seemed advisable have been made, at the cost of discrepancy with the succeeding tables. The general scheme of arrangement is admittedly devoid of pretension to ethnological order, since in the chaotic condition of the population at the present time an attempt to achieve such would inevitably be a failure. On the other hand, an arrangement purely according to the position occupied by the caste in social estimation would only be possible for each Province or State taken separately, as there is but little uniformity in this respect if we regard the circumstances of the coimtry as a whole. The general rule, therefore, has been to place the caste concerned in the group to which it is assigned in the part of the coiintry in which it most prevails, so as to get at least approximate accuracy. There are, however, obvious flaws in the grouping itself, due in great measure to the adoption of functional classification to an excessive extent. For instance, amongst professions we have the arts of singing, dancing, and acting, which in India are anything but reputable in public estimation. Again, in the large class of agriculturists are included the field labourers, who are. in many places but little more than agrestic serfs, and the forest tribes, who, though chiefly agricultural, no doubt, are lower in the scale than the village artisan. Thus, every one of the chief Classes must be taken by itself, and the relative positions of the groups in it considered as subject to this limitation. But for the purposes of the present review, a little more detail seems advisable in the combinations, and as the recognition of function in connection with caste has been already carried so far, there is no sufficient reason for seceding from it, on Luther's principle of pecca fortiter, if the principle be a wrong one. In the table on the next page, accordingly, the 60 groups of the original distribu- tion have been collected into 21 classes, and against each is shown the ratio borne by it to the total population dealt with in the return, which amounts to 286,900,000. Taking the return as a whole, the two points that are most prominent are, first, the indication given by the small proportion of the purely agricultural castes of the extent to which other classes join in cultivation, and then, the fact that nearly 16^ per cent, of the population is returned on other than functional considerations, thus showing the inutility of comparing this return with that of occupation. In fact, in all cases above those of the lowest, the members of the caste have diverged widely from the means of subsistence from which they respectively take their name, and the divergence is very often indicated by the establishment of a fresh subdivision, higher or lower, according to the occupation in question. In some of the provincial returns the several subdivisions have been compiled into an index, and in others the more important have been shown in the return subordinate to the main caste. This detail, however, was beyond the scheme adopted for the Imperial tables, which deal accordingly with the latter only. It is now proposed to consider Table XVIL— A., re-grouped as described above. The agricultural class then contains nearly 30 per cent, of the population under Agricultural three groups The first, that of the military section, includes those castes and gro^Ps- tribes which have risen to power at different stages in the history of their province. Amongst them the largest, that of the Bdjfut, is spread all over India, but to a far less extent to the south than in the continental portion. The title is an extremely comprehensive one, as must have been seen from what has been said regarding the A a 3 190 Table Proportion Proportion Class and Group. to 10,000 of Class and Group. to 10,000 of Population. Population. I.— Agricultural 2,988 VIII.— Traders 428 1. Military and Dominant 1,025 15. Traders and Shopkeeper^ 424 2. Other Cultivators 1,670 16. Pedlars 4 4. Field Labourers 293 IX.— Professionals 755 ''6. Priests 539 II. — Pastoral 583 7. Devotees, &c. 95 8. Temple Servants 11 3. Cattle Graziers' 29. Shepherds and Wool Weavers 403 180 9. Genealogists 10. Writers 21 89 X.-Arts, &c. 40 III.— 5. Forest Tribes - 551 11. Astrologers and Herbalists 11 12. Musicians and Ballad Reciters 22 13, Singers and Dancers, &o. 5 IV.— 36. Fishers, &c. - 288 14. Actors and Mimes 2 XI.— Carriers and Pack Animals 34 v.— Artisans 1,007 XII.— Vagrants 121 18. Goldsmiths 58 43. Grindstone Makers 1 20. Blacksmiths 91 44. Earth Workers, &c. 39 21. Carpenters and Masons 120 45. Knife Grinders - 1 22. Brass and Copper Smiths 11 46. Mat and Cane Workers 22 23. Tailors 26 47. Hunters and Fowlers 33 26. "Weavers and Dyers 327 48. Miscellaneous Vagrants 14 28. Cotton Cleaners 30 49. Jugglers and Acrobats 11 30. Oil Pressers 163 31. Potters 122 XIII.— Indefinite Indian Castes 107 32. Glass and Lac Workers 5 33. Salt and Lime. Workers 53 XIV.— Indian Christians ; 64 34. Goldsmiths' Refuse Cleaners — ^Q "NTntTVP 63 3.5. Gold Washers and Iron Smelters 1 59. Goanese, &c. 1 XV.— Musalmdhs hearing Foreign 1,197 Titles- VI.— Personal Services and Pood, &c. 488 XVI.— Himalayan Mongoloid 8 19. Barters 130 24. Grain Parchers, &c. 49 XVII.— Burmese and Chinese Mongoloids 254 25. Betel-leaf, &c. Sellers 8 52. Burmese, &c. 253 27. Washermen 98 53. Mixed Races 1 37. Domestic Service, &c. 15 38. Distillers and Toddy Drawers 39. Butchers 167 21 XVIII.— Western Asiatics 4 XIX.— Eurasians 3 VII.— Leather Workers and Lower Vil- 1,073 XX.— Europeans 6 lage Menials. 40. Leather Workers 488 XXI.— Africans 1 41. Watchmen and Village Menials 42. Scaveiiffer.s 446 139 Tptal 10,000 Class I. MilUary 1. Rajput 2. Jat 3. Gujar 4. Mai-iltliii 5. Babli:in G. ISTair 7. Kalla S. MaraAva 9. Vellaiua 10. Khandait 11. Awan 12. Katlii I.'}. Meci 11. Kodami extension of Bralimanism by fiction of comnaon origin, in which the warrior caste plays a leading part. Practically it is not the caste, but the tribe of Rajplit that is regarded in the present day, so the general term has fallen into the hands, more or less, of the lower grades of social aspirants, whose acquaintance with the intricacies of Kshairia genealogy is of the slightest. In the Jat, again, we find a caste which is so divided internally that the tribe has taken its place in social estimation. It is still more of a northern denizen than its predecessor, and is mainly found in the Panjab and the adjacent tracts of Rajputana and the upper Jamna. In fact, it has strayed but little from the tract it occupied on its entrance into 85,729,227 29,390,870 10,424,346 6,688,733 2,171,627 3,324,095 1,222,674 980,860 410,983 313,881 479,783 671,272 616,328 41,996 365,726 32,641 India from the north. The ethnology of the tribes that compose it is most interesting, as they contain better trace of their origin in the Caucasian region of 191 ancdent history than any oisher, but this is not the place to enter upon the subject. The GUjm- is another northei?n tribe, but more like the Rajptit than the Jat; it is composed of varied elements. In the Panjab it is mainly agricultural, though it tends towards cattle grazing in the southern plains. Elsewhere in India the title generally implies the latter occupation. Going further souths we meet the division of the Bombay; Presidency to which it gives its name, and from this point southwards the latter almost invariably implies a trader from Gujarath, without any reference to the northern origin of the caste. It is undoubtedly the relic of one of the later Skythian waves that flooded upper and western India, and many of the old forts and caves of the Satpura and north Deccan are attributed to rulers of this race. We then pass to the Mardtha, the ruling caste in the Deccan proper, who exceed the Giijar in numbers, but not in expansion. The number is swelled, no doubt, by the use of the term in contradistinction to a linguistic or sectarian community with which the Maratha speaker is brought into contract. Thus, wherever there are many Lingaiats, the north Deccani will call himself a Mar&tha, though a,mongst his own people he would be a Teli, Kunbi, or some inferior caste. The Bdbban is a caste confined, according to the returns, to Bihar, but the Bhuinhdr of the adjacent territory of the North-West Provinces should, no doubt, be added. It is an offshoot of a community of fairly pure Arya blood, descended from some of the earlier settlers of Hindustan. The Awdn is a Panjab caste, said to be one of the branches of the earliest Arya dispossessors of the Skythic races, and, like the fast of its compeers in that corner of India, its tribes are more cohesive than the caste as a whole. The Kdthi seems to be descended from the Kathaei who defended their country against Alexander, and who afterwards migrated, whether voluntarily or not is uncertain, to the south-west, where they give their name to "the province of Saurashtra. The Meo is an interesting caste of the. northern tracts of Eajputana and the south of the Panjab. It is mostly converted to Islam, but seems to have been originally a wild tribe of the Central Plateau. The EJianddit is the military caste of Orissa, and is said to be of the same stock as the Bbuiya, a Hill tribe of the central belt. From constant employment in military service, and intercourse with the warrior caste from Hindustan, who conquered this tract, they have made a decided advance in appearance and general civilisation since they took service with the foreigner in a strange land. We have now left the military classes of the Dravidian country, of which the Nair, who has been mentioned in a preceding part of this section, is the most numerous. The Kalla, or thieves, are, with the Marava, the ruling class of the extreme south-east. The former are the Colleries so often mentioned by Orme in his account of the British operations in these parts. The Vellama, or Tellama, for the initial is interchangeable all over the Deccan, consider themselves, from their title, autochthonous in Telingana and the Kanarese Deccan. It is a question, and an interesting one, whether these frequent instances in India of caste names meaning " those of the soil " signify title to the possession of the land or those bound down to it. Examples of both usages can be found. Presumption seems rather in favour of the interpretation that these classes represent the pre- Aryan inhabitants who were not driven off the land in the Gangetic basin and western India, but were ascribed to it as serfs in the plains, and left as occupants in tracts which did not tempt the foreigner. In the case of the YelJama, probably they were fixed in possession long before the stray Brahman began to thread his way through the Kaimur or Vindhia. The last name that requires comment is that of the inhabitants of Ooorg, the Kodagu, a peculiar and warlike race, secluded from much of the influences that have caused the general fissiparity of the Dravidian population of the plains. Group 2 of this class is, necessarily, the largest and perhaps the most miscel- ' laneous of all, though in the latter respect . the group of Musalman who own to foreign ' titles approaches it. First of the castes ' that are here included comes that of the , Kunbi or Kd/rmi. The former designation , is that of the Deccan and west of India ; the latter is used in the Gangetic valley. Both mean no more than tiller of the soil, and are therefore minutely subdivided. The Kurmi is returned from the North- West Provinces and Bihdr chiefly, but he is found, too, in the north of the Central Provinces. Closely allied to this caste, in fact, in some A a 4 2. Cultivators 47,937,361 1. Kunbi, &c. - - - 10,631,300 2. Mali 1,876,211 3. Lodha 1,674,098 4. Kachhia 1,384,222 5. Koeri 1,735,431 6. Vellala 2,254,073 7. Vakkiliga 1,360,558 8. Lingaiat - - - 655,491 9. Panchamsali 482,763 10. Kaibartta 2,298,824 11. Nama Shadra 1,948,658 12. Kochh 2,364,365 13. K61i 3,058,166 14. Eeddi 2,665,399 , 192 parts of the country, almost subdivisions of it, are the Kdchhia, and Kunjra, market gardeners, and tlie Koeri, the Lodha, and the Mali, cultivators. All but the^Mdli are chiefly found in Hindustan and ' its neighbourhood. The Mali is so bound up with > the Kiinbi in the Deccan that the ordinary term for an agriculturist is the combina- tion of the two, Kunbimali. We find names of the same comprehensive character amongst the cultivators ,of the south, as, for instance, Velldla, the Tamil equivalent for the Kunbi class, and the Walchaliga, which corresponds to it in Mysore. The Lingamt, too, of which something was said above in connection with sect, is shown without distinction, and the Panchamsdli, which is one of the terms included in the former, appears independently to nearly the same extent as the parent. Here we begin to find discrepancy of function, for the Panchamsali must, from his name, have once been a weaver, or sprung from the village menial class, whereas he appears here as a cultivator and generally is one. Again, to cross India, the Kaibartta, the great agricultural caste of Bengal is, or was, entirely devoted to fishing, and it is even stated that his name is derived from his acquatic tendencies, but he is now rooted to dry land for his main livelihood, whilst the Bihar caste, that is said to have once been a branch of the Kaibartta, has changed its name to K^wat, and will be found amongst the fishers. Take the Ndmashudra,, again, the lowest caste, or very nearly so, in eastern Bengal. Here they are cultivators, probably because there has been no ruling class from the west to oust them from the soil, and the proprietary right belongs in great measure to non-agriculturists, who require their property to be worked by, other classes. But in Assam they are fishers. Amongst the more interest- ing castes in this group are the Kochh, originally a Mongoloid tribe that came down the Brahmaputra valley. They are now completely Brahmanised in title and certain portions of their domestic ceremonial, and have abandoned Assam for the submontane tracts of Bengal. The Koli, who is returned in the largest numbers from the "Western Presidency and the States of Haidrabad and Central India is not a homogeneous caste by any means. The philological connection between this appellation and that of K61, and similar terms in, more eastern parts of the continent, has not been established, but there is evidence, nevertheless, of the pre-Aryan existence of a race corresponding to it. In Gujarath, where it forms the bulk of the lower cultivating community, it is called " the local," though there are several endogamous subdivisions, chiefly territorial. Along the Grhats there are other Kdli of an entirely difierent community in the present day but of the same stock. On the coast, again, the K61i represents the early fishing interest, like the Kaibartta on the other side of India. Everywhere he is dark, and of the same physique as the Hill tribes of the Central Belt, except on the plains of Gujarath, where the mixture with the higher races from the north seem to have improved his appearance as well as his intelligence. It is a question how far the K6ri, or low class weaver of the north, is allied to the K61i of Central and Western India. Btymologically the titles are the same, but socially the former is an offshoot of the village menial of impure habits in connection with leather, whereas the Koli does not by his touch pollute the twice-born. But ao far as the census is in question, the K6ri is classed with the weavers, whilst all the others appear under the present head. The only other large caste that is not purely provincial in its dispersion is the Beddi, or Kapu, the great cultivating caste of Telingana. The third group of agriculturists is 3. Field Labourers 8,407,996 Miisahar 622,034 ]iagdi - 804,960 Bawiiri 612,430 Diibla - 172,052 Pale 2,242,499 Palla - 814,989 Mala ; 1,366,520 Cheruma 523,744 i tradition of the Fairs is fa trly trustwo anything but well-defined, as the boundaries between it and the lower grades of land- holders and Hill tribes on the one side, and of village menials on the other, depend simply on local conditions. In the open country the position seems to be, generally speaking, the lower, and rises as the tract falls in futility or ease of access. The bulk of the group is found in the Madras Presidency, where the agrestic serf is an actual though unrecognised institution along the Malabar coast.* It is probable that the * Camoens mentions the distinction — " A nohre Naires chamados sao ; e a menos dina Poleas tern per nome ; a quern obris;a A lei nao misturar a caista aintiqua." 193 Class II.— Pastoral Group 1. Cattle graziers, SfC- 1. Ahir 2. Gauli, Goala, &c. - - 16,721,494 11,569,319 8,155,219 2,237,323 nrt^3w h °^^r' «"«^/« the Pulya, who are of an inferior race, and still Ct ?l r°^.. 'S the estates on which they labour. In other parts of Southern vSL L^ 1 V v'S^'' *> corresponding castes do not belong to the group of '^^S T^ V''''^^''\^%^r^ ""'^ ^^"^^^^"^ i" *^« neighbouring provinces and wViI%.« V. ■ '^'^!' ^\ difference between the PdU, the Mala and the Paraiyan, ,^ oli fi^f Imgnistic only, does not seem to quite justify the separation of the two m classification as was done on the above analogy, in default of local knowledge.* m the west of the Bombay Presidency there are two distinctly agrestic classes of labourers confined to one or two districts. These are the Bulla ^J the DUdia, who were ascribed to the soil by a class of agricultural Brahmans who colonised the southern portion of Gujarath. It is curious, by the way, to note that where Brahmans take to. colonisation there are always predial serfs attached to the estates, as in the tract now in question, on the Malabar coast and in Kanara and in Orissa. In the Nortn-West Provinces and Bihar we find the agricultural labourer but little reclaimed from the tribes of the lower hills of the central belt. The Bdqdi and Baioari and Mmahar are all instances of this, and with the Eajwar, a smaller caste of the same sort, are all related to the Bhuiya, Saharia, Chero, and Kh^ro of Class III. f rr^" J ^ village watchman, such as Dhakar in the west and Dhanuk in the centre ot Hmdostau, is confounded partly with these functionaries, as he is in Madras. The next class in the list is that of the castes accredited to pastoral occupations, of which the group engaged with horned cattle is the most numerously represented. The Ahir, under this title, is chiefly found in North- Western India, where he has, in his time, given more than one dynasty to local thrones. To him is attributed a Skythian origin, and there is no doubt that the Ahir once prevailed a little way south of the Vindhia range, and was in the Indus Valley at a very early period in the history of India. The largest subdivision of the caste, known as the Goalwansh, extends from Oudh far into Bihar. The next caste mentioned in the margin is probably only functionally connected with it, though the etymology of the title is the same. The Qodla, Gauli, or Golla, is the Ahir of the south and east, and spreads all over Bengal and the peninsula. The second group in this class comprises those castes who specially attend to sheep and goats, and who also are engaged in the preparation of wool and the coarse blankets universally worn in wet or cold weather. In the north of India the Gadaria represent this class and the Dhangar of the Deccan is nearly akin to them. The Hatgar is a subdivision of the Dhangar, but more specially devoted to weaving and the spin- ning of black yarn from the fleeces of their charge. The Kuruhar are the shepherds of the South Deccan and Telingana generally, whilst the Idaiya and Kuriimba are the Tamil, or South Dravidian, representatives of the group. Of the larger provinces, Bengal and Burma have the smallest proportion of this class, owing to the com- parative scarcity of grazing land in the former and the damp in both. The latter reason, with the great amount of forest, probably accounts for the absence of the shepherd from Assam. In Western India, Kajputana and Grujarath, the Eabari and Bharwad are the chief shepherd classes. Nearly all are, from the nature of their calling', nomad for most of the year, though some of the larger clans keep up connection with some headquarter village or district. It is impossible to enter into the detail of the heterogeneous class of i'orest Tribes ; so, for convenience of review, the chief items have been collected into territorial groups, several of which denote, also, racial distinctions. The marginal table over page shows the main results of this arrangement. The largest group is that of the tribes of the north- eastern hills of the Central Belt, amongst whom, however, we must include the detached Group 2. Shepherds, &c, *- 5,152,175 1. Gadaria - 1,294,830 2. Dhangar 1,306,583 3. Kiirnbar 1,059,185 4. Idajga 665,232 5. Eabari 434,788 6. Bharwad 128,271 * The caste chapters of the Superintendent's Report have not yet been received.- X 78388. B b -S. A, B. 194 colony of tlie Santhals, further north. }las s III. Forest Tribes 15,806,914 " 1. Siinthal 1,494,045 2. K61 ^ - 474,969 3. Korwa 158,700 4. Ho 150,262 <^. 5. Muiiila 410,624 6. Bhuiya, &c. , 909,822 7. Pan 341,740 8. Kliarwar 112,298 9. Oraon 523,258 flO. Baiga 136,478 < 11. Kiir t.12. GSiid 155,831 3,061,680 'J 3. Bhil j 14. Bhilala 1,665,474 175,329 "] 15. Kirar _16. Menu 175,508 669,785 , '17. Kathodi 77,705 _ 18. ffarZi 168,631 <; 19. GMt-ThdMr . -. ; 130,481 20. Naikada 74,479 J2\. Dhanlia 67,451 '22. Kaudh (Kbond) 627,388 23. Sawara 438,317 24. Halaba 102,643 <^ 25. Gadaba 34,127 26. Yanadi 84,988 _ 27. Jatdpu 81,152 r28. Tdda - . ' V 739 { 29. Kota ' 1,201 [30. Irula . - 58,503 f31. Khasi 172,150 1 32. Kaohari 243,378 < 38. Garo 150,227 34. Mech 96,873 35'. Tippera ' 99,395 \ 36. Mikir 94,829 \ 37. Naga; - , , - 101,568 (_38. Chihg-pau, &c. 3,483 '39. Kiiki' -; . - 25,940 40. Lushai 43,840 41. Kathe (Manipuj-i) "> 42. Khyin {Unspecified) 84,540 82,710 43. Khyin Khwe-ini ^44. Khyiu Mrd 14,200 15,666 the race on to The whole group may be conveAtioixally termed Kolarian, as the language o^f all bpt the two last belongs to that family. But they are, nevertheless, of the same ethnic stock as the bulk of the next three groups. They furnish the bulk of the labourers wl^o migrate into Assam for a term of service on the tea estates, &c., and in Easter]|i and Central Bengal, too, are found in consider- able numbers as general labourers, wherever strength and endurance are required. The discrimination noted in the margin. between the Ho, Munda, Kol, and Korwd, is by no means accurate. , The term Kdl, for example, seems to have been applied to certain , tribes by outsiders, as the people concerned recog- nise no name but that of H6, or Miind^. The Bhuiya, again, ie not a homogeneous body, but includes Bhumij, Bhinjwa, Bhiinjia, &c., and the term Baiga, which is that of a tribe in the Centrar Provinces, is applied by some people in Ohutia , Nagpur simply to the magician of the community, possibly because he belonged to a tribe preceding, in occupation of the place, that which now employs him in the above capacity. The second group apper- tains more especially to the central ranges of the Hill Belt, the Bhil, Bhilala, and Kurhu being found to the westwards, and the Baiga in the middle. The Gdnd, who, as shown in the language section of this chapter, gave their name to the whole tract, are still spread all over it, north and . south. They do not come much .within the Chiitia Nagpur districts, Berar, Haidrabad, or North Madras, though all these tracts con- tain certain of their tribes, but are chiefly found in the Centrg,! Provinces and the Orissa hills. The subdivisions are very numerous, and in some cases amount to a completely separated community. The Kv/r spread from Chutia Ndgpiir, across the south of Central India into the hills which bound Berar on, the north. The Bhils there the western extremity of the Satpura and now " -- meet them, and carry Vindhia. Here they seem to have founded the now separate tribes of Bhildla, Pawada, Dhanka, Naikada, and Tadwi, of whom the third and fourth are the Grujarath representatives of the great tribe, and the last has become Musalmdn. The Kirdr is almost confined to the hills where the Central Provinces, Central India, and the North-West Provinces meet. It is doubtful whether it is an independent tribe or an ofishoot from one of the larger ones. The Mena, or Mina, are the originals of the Meo, who have moved into the northern plains adjoining the Jamna and have embraced Islam. From their title it might be thought that, like the Kaibartta, they were connected with fishing, either by profession or by tribal symbol, but their gods are those of the landholding classes, and there is no tradition of other occupation than those of cultivation and robbery, for they formerly kept the whole country round their bettlements in terror of marauding expeditions. It is doubtful whether they are properly classed as forest tribes, and whether they should not rather be placed amongst the Skythian wanderers, like the Ahir or Kathi. In the next group, Nos. 17 to 21, the two last have been mentioned above as offshoots of the Bhil and the three others are confined to the country immediately above and below the Ghat range. The Kathode,, or Katkari, get their name from the occupation of pre- paring catechu from the tree, to which they were originally addicted. The Wdrli 195 are hill cultivators of a • low type, and the Thahurs, a branch of the same race, specially distinguished , by this,' honorific designation, and in position and habits somewhat above the others. We then cross to the Orissa hills and their western extension. Here we have the Kandh, or Kh6nd, of whom much has been written, and the Sdwara, mentioned, it is said, by the Greek geographers, both being confined .to the eastern ranges- The ffalaba, probably allied to the G-5nds, succeed them towards the west, and the Qadaha are also adjacent to the south-west. The Yanddi are the tribes of almost detached masses of hills in Telingana, like the still smaller tribe of the Chentsu. The Nilgiri group of hills is represented by the Toda and .iKo^a, who • have been already mentioned as small or local tribes, curiously isolated •from the rest of the Dravidians, and by the Irulas, who spread down the slopes on the -Mysore side. There are other tribes of much the same class occupying the Ghat forests, Malabar, and the Annamallai and other groups of hills in the extreme south, but they are of small numerical 'extent, and have not had the advantages in the way of advertisement enjoyed by the Tdda. The three remaining groups are Mongoloid in race, and chiefly Assamese in habitation. The Khdsi, though linguistically distinct, are racially allied, in all probability, to their neighbours in the hills east and west of them. Tbe rest belong to the;gKeat -Bddo group of the Assam valley and the hills -to the south thereof. The Tippera, Ai is true, is now separated from the parent stock by sundry tribes ot Zho and Kuki,' hut its language shows that the separation is of comparatively recent date. The small tribes of the Abor and Mishmi have not been specially mentioned, because,. like their western neighbours, the Akka and Dafla, the majority of the community dwells, outside the British frontier, and the population found within the red line consisted partly of offshoots settled there within the few last generations, or casual families or gangs come down to market. The Naga, Mikir, andSingphd, or Kakhyin, have been mentioned in connection with their respective languages. The first is the subject of an interesting memorandum by Mr. Davies, appended to the Assam Census Report. The Burmese tribes belonging to the last- named family have likewise yielded a corresponding contribution to the Report on the Census of Burma. As regards the last group, very little, as has been stated above, is at present known. The Manipuri, however, has his tradition of ancestry from the demi-god Arjun, and Bindraban, the i scene of the youthful frolics of Krishna, is the goal of his aspirations, as Benares is to the devout Khas and Gurkha of Nipal, who are more Kshatria than the Kshatrias of the plains. But the mass of the population of iManipur is Mongoloid, eveniin .title, and Brahmanic proclivities are confined to the court aad its entourage. The forest tribes of Burma,. when not returned as Khyin or Kakhyin, have been absorbed into the general titles of Burmese or Karen. The class of fishers is broken up into a large number of small castes. Those given in the margin are the best represen- tatives, either of the group as a whole or of a special province, but they are probably mixed up together in the returns. For instance, the Dhimar and Jhinwar are pro- bably the Panmb and Central Indian equi- valents of the Kahar, from, whose ranks they sprang. Looking at the peculiar position occupied everywhere by the class of fishers, especially those who are very largely en- gaged in non-piscatorial work, it is clear that there is some racial basis for the anomalous rules and .privileges attached to this caste. The fisher will carry any •burden that is supported on a pole across his shoulder, but may refuse a head load or knapsack. He can give water to men of far higher c^ste, and become a servant inthe recesses of their establishment ; he can even knead dough, &c. for their bread, but must not lay a finger on it once it has been put into an oven or on the bakmg- nlate Yet he is allowed to indulge in ardent spirits and to eat strange food on which the ortKodox may not even look, and is admitted to functions from which far more wealthy and reputable castes are rigorously excluded- As regards the nomenclature of the castes selecied for mention, the ^aMr is the form adopted throughout the ISTortli-West Provinces and Bihar; the Jhmwar is tlie Pan]ab equivalent, and the Tfhl'nu^r'l'ha.t adopted in the Central Provinces and its" neighbourhood. In Assam we have seen that the two castes, which are Qlassed. in Bengal as agricultural cpnstitute, S the Kewat, 'the bulk of the fishing class, whilst the Kewat of Hindustan Is Bb 2 Class iv. Fishers 8,261,878 1, Kahar 1,943,155 2. Mallah.. 1,1^7^,544 3. kewat , 989,352 '4. Bhoi IS06,190 5. Jhinwar 489,819 i 6. Dhimar 287,436 7. jGaurhi . - 317,111 8. Machhi 260,496 M&S \ 196 probably the Kaibarfcta of tbe Lower Granges. The Maliah is prevalent ia both these provinces. In the table, some of this caste are attributed to Haidrabad, but it seems very likely that they belong to the agricultural labourers, called Mala in the Madras Presidency. Amongst other "widespread castes of this group comes the Bh6i, which is apparently the Telingana version of Ka,har, and is spread also over the Western Presidency and Rajputana, and the Besta, which prevails still further south. The rest are almost, entirely local. Muhano is the Sindh fisher caste, Maohhi is only found on the Indus and in Grujarath, and Sembadawan, Moger, Palle, &c. appertain to the south of the peninsula, east and west respectively. The functional classes contain little but linguistic or territorial groups, which only need comment in special cases, or where there may have been confusion in nomen- clature or tabulation. The first four groups are taken; together, because, in popular esti- mation, they are the special care of Viswa- karma, the Hephaestus of the later Yedic theogony, and in the south of India are almost invariably taken as constituting a single caste, though with five non-endo- gamous subdivisions. Elsewhere, however, the latter only are recognised, and the five vary respectively iu rank according to locality. The function of the gold and silversmith has been described in an earlier chapter. His aid in making and re- making ornaments is constantly in request where capital invariably takes this form of investment. Owing to the majority of the members of all five groups being returned in Madras under the general title of " the five arts," the number there has been distributed according to that of the portion of each class which was separately returned, so that the division is partly arbitrary. The three castes shown in the first group are simply the general title, with its Kanarese and Tamil equivalents, and a good deal of the second has doubtless passed into the first. The second group, again, consists of two linguistic groups. The first, Luhar, is a general term, though, as explained above, the Luhar of the Panjab will decline connubivm with him of the Deccan, as will the Kammar of Madras with his namesake of Bengal. The carpenters, again, comprise six groups, none of which, save for differences in language and locality, are mutually exclusive. The original carpenter caste was the Sutar, or Sutradhar, but the Barhai, from vardh, to cut, is equally Sanskritic, as is Kathi, Kbati, from either Kdth, timber, or EMti, a wooden bedstead, &c. Badagi and Asari are the Dra vidian versions. The mason is poorly represented. In the Deccan he forms a separate caste, but the Eaj of Upper India is probably more comprehensive, and includes all who work at masonry. In some provinces or tracts, of course, his services are scarcely required by the great bulk of the population, and in all, the carpenter undertakes most of the building, aided by unskilled labour. The brass-smith is growing annually in wealth and importance, but his functions are not yet so specialised as those of the carpenter or blacksmith. Of the two main casteH mentioned, the first carries on the whole business in most parts of the peninsula, and sells what he makes. In the Panjab, however, the Thathera" sells and the Kasera makes, and in the North- West Provinces the latter moulds and the former polishes the vessels and platters, &c. There is a curious affinity between brass- ware and Brahmanism. The great centres of brasswork in India are also the Brahmanic centres of pilgrimage, such as Benares Nasik, and Madura, or political centres where Brahmanic influence is supreme, as Poena and Tanjore. Then, again, in all the temporary bazdrs that are set up during the time of pilgrimage at shrines and bathing-places, the booths of the seller of brass vessels, images, lamps, and so on, are perhaps the most conspicuous of all in number and variety, for the commercial element is as prominent at these gatherings as the sacerdotal. It is, therefore, plain t];j§i,ti the production of these wares must be carried on by a greafmany paore than the r^l^tively insignificant number set forth in. the return. Class V. Artisans 1 28,882,551 Group 1. Goldsmiths, SfC. 1,661,088 1. Sonar . . - 2. Aksale 3. Tattan 1,178,795 307,670 .56,044 Geoup 2. Blacksmiths 2,625,103 1. Luliar 2. Kammar 1,869,273 666,887 Group 3. Carpenters and Masons 3,442,201 < \ '1. Sutar 2. Barhai 3. Tarkhaii 4. Badagi 5. Asari 6. Khati ~7. Gaundia, &c. 8. Kaj, &c. 681,790 932,718 696,781 452,339 100,409 301,476 76,995 19,770 Group 4. Brass and Coppersmiths 3oi,5i9 1. Kasera, &c. 2. Tathera 3. Bogar 161,596 60,837 37,002 197 Group S. Tailors 735,548 1 . Dai-zi and Shimpi 710,092 Group 6. Weavers and Dyers - 9,369,902 I. Julaha 2,660,159 2. Kori 1,187,613 3. Tanta 483,942 4. JQgi 424,219 5. Balai 305,635 6. Sali 394,064 7. Kaikola 316,620 8. Tatwa 328,778 9, Koshti 225,019 10. Ganda 291,768 H. Patniili 96,443 12. Khatri - 116,880 13. Eangrez 187,698 Group 7. Cotton Cleaners 85g,288 1. Pinjari 753,675 2. Kadhera, &c. 105,613 The next section of the class under review is that of the Workers in textile fabrics; This, as has been said in a former chapter, is an occupation evolved out of the dregs of the village population into comparative respectability, and in the case of "certain articles, such as silk, gold braid, and coloured goods, into a decidedly good position. But lihe mass of the castes are still not far above the village menial. The Kori, for example, is often returned as a Chamar or leather worker, to which tribe he once belonged. The workers in hemp and in coarse, undyed cotton cloth are elsewhere set far below the ordinary toiler at the loom. The latter is said to be occasionally ambitious of rising in the social scale, but circumstances are against all but perhaps the silk-weavers, whose wares are less affected by foreign competition. The castes mentioned in the margin are mostly provincial and linguistic, except the three last, of which the first represents the high-class weaver of silk, &c., the second the Gujarathi weaver of the finer class of wearing apparel, froin which the former sprang, and the last is the dyer, who, like the Jolaha, is both Brahmanic and Musalman, sometimes even according to the material he works in. The Darsi, according to his modern and Persianised appel- lation, or the Shimpi, in the older version, is somewhat of the rank of the middl^-telass weaver, and is developed from the calenderer, Ohhipgar or Bhausar caste, which, again, is allied to the weaver. It is more or less of an urban growth, however, and is not mueh found in villages. The cotton-cleaner is usually a Musalman, and is found all over the country, wherever cotton is grown and weavers work in it. The general name for the caste is Pinjari, or Penja, but the Musalman improves it into the Persian title Nadaf, or Dhunia, and the Brahmanic cleaner in Hindustan and the Central Provinces calls himself Kadhera. In Bengal there are several local titles in both religions. The two next groups, those of the oil pressers and the potters, are very simple, the different items representing simply lin- guistic variations of the same functional group. The Teli is the r general title. Ghanchi corresponds to it in Gujar^th, and Ghaniga in the South Deccan. Vaniya is the Tamil equivalent, and Kalu is a Bengal subdivision, based on some in- feriority of process in extracting the oil. As to the potters, the only variation from the general caste is in the shape of a few who have returned their function as brick- making, which in some parts of the country The second item is merely the southern title of Group 8. Oil pressers 1. Teli and Ghanchi 2. Ghaniga 3. Vaniya 4. Kalu Group 9. Potters Kumbhar Ktisuvan 4,672,907 4,147,803 142,374 186,297 191,395 •3,497>'5o6 3,346,488 138,097 has been differentiated from the rest. the potter. The glass workers are mainly makers of beads and bangles, in lac is very closely Group 10. G Workers. 1. Chtiiihar 2. Lahera Group 11. Sc Workers. 1. 2. 3. 4- Liinia Uppar • Agri ^ Rehgar Glass and Lac and Lime i55,oo3 55,618 32,139 i,53i,i3o 796,080 267,715 241,336 77,856. The class that work lUied to them in rank, but are not so largely recruited from Islam. There are numbers of small castes returned under special or local titles which come under the general head of Kachdri. Those mentioned in the margin are chiefly in. Bengal and Hindustan. The salt and lime workers belong to castes nearly allied to each other. .The iwrna are the prevailing caste of this group in Northern India. The Agri are found along the west qoast. Bb 3 l98 Group 12. Goldsmiths' Refuse ■6,2)6c> Cleaners. , 1. Niaria 5^808 2. Jhdlgai- 555 Group 18. Irmi Smelters and 24,8q3 Gold^'Washers. .-.-, 1. Jhora 7;837 2. I)}iang9,ri 3,673.'. 3. Asuri 3,552 4, Deoli 2,289 The JJjapdr are the Dravidian equivalent to the Ldnia. The Behgdr are confined to Rajputana, where the S^mbhar Lake gives employment to a considerable number of this class. The next group, that repre- sented by the Niaria and Jhdlgar, two names for the same small community, busies itself with recovering the chips and filings of precious metal from the refuse of gold- smiths' workshops. The iron smelters are ■found in some of the Hill tracts of the Panj^b, the 'Central Belt, and the Ghats of Bombay. They are a small and scattered community. The gold washers, 'too, are very low in both numbers and rank. Most provinces have a stream or two, in the bed of which minute portions of gold are found mingled with the- sand, and by careful panning the outturn is just enough to support the family of the operator. The heavy mining work in Mysore is largely in the hands of EuropBan labourersJ - The position of the two first groups of the class of personal servants have been described in a previous chapter. The Nai, N"apit, Nhavi, &c., besides shaving, which is a ceremonial of considerable importance, acts as formal messenger and go-between in cases of betrothals, &c., and is also a leech of em- pirical skill. The Hajdm is the same caste under its-^ Musalman title. Ambattan' and Mangala are found in the south and south- east of the peninsula, and Bhandari in the Central Provinces. The Dhohi shares the low rank of the potter, partly owiiig to his use of the donkey in his profession. The second name in the group is that used in the Kanarese and Telugu country. The two next appertain to further south, and the last is purely Maratha. The group of other servants is a miscellaneous one, many of the items in which are scarcely entitled to the name of caste, as they are not all endogamous or otherwise distinct from each other socially. The first on th« list. Gold, is the rice pounder of Western India. The second, the wat^r bearer, is, returned as a separate caste chiefly in Hindustan, and enjoys the name of "Paradisaic" from the boon he carries. The other two are simply hereditary domestic servants of the courts of the chieftains of Rajputdna and Central India, some of them, no doubt, blood relations of their master. Most) of the fourth group are connected with the great fisher caste of the Kahar, or some corresponding division. They are most strongly represented under the titles selected in Hindustan and Bengal. The first-named, Bhadbhtoj a, is found, also, far to the south and west of those provinces, but usually, asserts an origin in the, north or along tbe great rivers. The next group is very undermanned, considering the ex- tensive and, in fact, universal consumption of the articles in which they deal. But in many parts of the country the sellers are also; growers of the leaf, and thus appear under the head of agriculturists. The Tdrnboli is a distinct caste in Central India and the Ndrth Deccan, but in Hindustan and Bihar it is probably a sub-caste of the Barai a large cultivating class, affiliated in some parts, of the country to the Kachhia, and tlius nnaliy thrown back on the great Kiirmi stock. ' ' The class that traditionally lives by toddy drawing and cultivating the palm tree is a very large and varied Oite." Th^'items selected, too, are somewhat peculiar in not Class VI. Persdiial and Domestic Service, prepawtion of Eood, &c. 1 ■'■■■» 14,019,626 Group 1. Barbers j 3,729,934 ]., Nai, &c. - ■ ' ' 2. ' Hajam 3. Ambattan 4. ', Mangala 5. Bhandari i 2,532,-067 605,721 186,187 154,438 103,026 . Group 2. Washermen 2,824,451 1. Dhobi , - , - 2. Agase 3. Vannan ' - 4. , Sakala 6, Parit 2,039,743 126,710 - 258,508 327,720 60,129 Group 3. Service, Sfc. 430,065 1. Gold 2. Bihisti, &c. 3. Khas 4. Ohakar 33,804 98,824 215,200 25,706- Group 4. Grain Parchers and Confectioner's: 1,407,169 1. Bhadbhunja 2. Kandoi 3. Halvai 4. Giiria, &c. '343,308 524,155 260,801 141,628 Group 5. Betel Leaf Seller^ and Perfumers. 236,5o7 \. Tamboli r,- 222,048 199 being, in most cases, linguistic largest GrKOUP 6. Distillers and, Tdddy 4,785,210 Drawers. 1. KahU 1,195,097 2. Shaha {Si'uiri) 525,698 3. Shan'a 690,434 - 4. Tiya 53«,075 6. Bhandari 170,014 ■ 6. Idga .196,901 7. Iluva - 703,215 8. Q-amalla ■ 122,322 9. Graundla 235,902 Group 7. Bnuhei-s 605,890 1. Kasai 302,612 2. Khattk 293,771 variations of the same caste. By far the number, of the communities belong to the coasts of Madras. The Bhandari prevails on the Bombay coast, and the 8Mha, or Sunri, in Assam and Orissa, whilst the Gaundla is returned only from ■ Haidrab^d. The Kaldl, or Kalwdr, is the general distiller of spirits, as distinguished from those who deal with' or in toddy. He is found in most provinces, and in Upper India has furnished one thl-oue with a dynasty. On the Westerti Presidency his place is taken by the Parsi, but above the Grhats he begins soon to settle down to his traditional calling and to thrive on it. The butchers are' of two classes, ^rahmanic, Kkatik, and Musulmahj Kasdi. They are most numerous, of course, amongst the martial tribes of the north and elsewhere in places where Musalmans dongregate. Some, however, the Hindus, for instance, do not in the present day kill cattle,, biit deal in the flesh of sheep and goats only. The class of leather workers and lower village menials is, next to the agricul- turists, to whom they are subordinate, the most numerous body in India, and includes all the tribes of the darker race, who pre- ferred a foreign yoke to exile from the tract they formerly possessed. The first , group comprises all who avowedly work in leather and other exumtB.oi defunct cattle, but the first and third castes, at all events, combine agricultural and village labour with their traditional function to such an extent that they mingle greatly with the group which follows them on the list. The Mochi is dis- tinctly an artisan, mpstly found in towns, where his occupation is specialised. In villages there , are distinctions recognised. Some will mend shoes, but not make them ; others, devote themselves to making the leather bags in which water for irrigation is "dlrawn from the well. A third only manufactures leather pots in which to store clarified butter, and so on. The fourth item, Saldlia, is the c^ste known as Ohucklers to Ahglo-Indiahs in Madras. The Madiga is a Kanarese and Telugu ca^te of the same nature, combining tanning with shoe-making, ^,rid the Bambhi is the specially Eajputana community of the same description. , From very early Brahmanic times the leather worker has been a member of the village community, p.nd always in, hi^ present degraded rank. It would, appear ,from AristophaneSj that it was much the same in Athens, where the hemp weaver, the Mang, as it were, had to, give place to the Cham^r, who, in turn, was to be dispossessed by a stilllower caste, the Chiihra, or pig-breeding scavengei;, before the degradation of the State was complete. The next group, that of watchmen, &c., is a trifle, higher, and several of the castes named, such as the Pdsi, Dh4nfi,h Birdd,. Mutrdsa, and BdmOsi occupy' a more reputable place in public estimation than tjie r^st. , The MaUr ' however, of whom the Dhed is the northern and the Eoldr the Kanarese variety, 'and the Mdng, are practically in the same ranl^ as OAawrtrs and, oth^r, leather workers, as they niostlv serve as tanners, thaugh they do not work up the hides they prepa're.' Their enterprise, in , going far a-field froih their native yillage for work, Glass Vii. Leather Workers and 30,796,703 the Lower Village Menials. Gkodp 1. Leather TVbrkers - 14,003,100 1. Chamar 11,258,105 2. Mochi , 961,133 3. Madiga 927,^39 4. Sakilia 445,366 5. Bamblii 220,596 GrKOOP 2. JVatchmtn and i2,8o8,3oo' other Menials. 1. Cosadh 1,284,126 2. Ghatwal 167,089 3. PAsi - 1;378,344- 4. Arakb 85,522 5. Dhanuk 883,278 6. Mehra ^ .- . - 226,216 ' 7. Malidi- 2,960,568 ' 8. Dhed ' - 508^310 9. Holar; 880,441 10. Mang , 690,458 11. 'Berad 659,863 12. Eamosi- 63,991 13. Mutrasa -^ 296,743 14. Paraiyg, (^Pariah) . - 2,210,988 Gjtoi'P 3. Scapengers and 3,984,303 Miscellaneous. 1. Mehtar 727,985 2. Chuhra' 1,243,370 3. Megh 148,210 *4. Bhuinmali 231,429 5. Bhuinhari 316,787 6. Dqm (Dumna) 1,257,826 200 and the demand for hides and skins for export to Europe, together with a strong local demand for tanners, &c. in some parts of India, where, as has been already remarked, European capital has been invested in this industry, tends to raise the- position of this class above that of the ordinary village menial, and as money is acquired, the demands for recognition as of a higher caste become more worthy of consideration, elementary schools are started, and the marriage .field is restricted, and in a few generations a totally new social stratum is firmly established, based, of course, on one of the heroes of the Brahmanic epics. As regards the territorial designations quoted in the marginal table, the Busadh is found chiefly, in Bihar and parts of Hindustan. The BhdnuJc extends into Rajputana and the Panjab, but it is chiefly in Bihar. The Pdsi is a •watchman, &c. in the North- West Provinces, as a rule, and more of a toddy drawer across the frontier, and the J.raM is his kinsman. Like the £eracZs, or Bedar of the South Deccan, they were originally hunters and fowlers. The latter are well known for their prowess under the Mysore Sultans, with whom they enlisted in considerable numbers. Their head-quarters were at Shorapiir in Haidrabad territory, the downfall of the Chief of which State is told by Colonel Meadows Taylor, in his autobiography. The Bdmosi is a small caste of the Deccan proper, not far removed from the K61i, but of more southern origin. , The Mutrdsa is the village watchman proper of Telingana. As regards the Paraiya, or Pariah, the well-known Tamil caste of labourers, it has before been remarked that his position is generally that of an agricultural labourer, but under special restrictions as to function and residence, like the Mahar of the Deccan. He has, however, no prejudices to overcome, and makes himself at home in any miscellaneous employment ; so, like the Oharaar of the north, his sphere is extending, though not proportionately to his numbers. As Brahmanic rules are more widely difi'used amongst the peaceful Dravidians, who received them later, than amongst the more varied and frequently disturbed commonalty of Hindustan, where they originated, the opposition to Paraiyan exaltation is m^ore marked. The third group, again, is composed of the lowest grades of the pre- Aryan classes, though the last, the D6ni, is of a very varied constitution, and rises to occupations not attained by the rest. The Mehtar, or Bhangi, is found under that title in Hindustan and Western India. In the Panjab the Chuhra takes his place. Both these castes are of fine physique, and very dififerent in appearance and stamina from the Dravidian type of the south. The Mdgh, or Meghwdl, is also found in the Panjab and Sindh, and is said to be of a still more northern origin, as he is accredited by Sir A. Cunningham to a Skythian parentage. The Bhainhdri, or Bhummdli, represent the Bengal and Assam scavenger. Their name clearly denotes the same connection with the land as that of Bhiimio, which is given to the corresponding class in Gujarath. In the Dam, with its varieties, we have a far more interesting community than that of mere scavengers. In the first place, it is a generic title, the derivation of which is uncertain ; but the term must be one of considerable antiquity, as it is used^ in the " Raja Tarangini," or " Annals of Kashmer," one of the early neo-Brahmanic produc- tions. There are one or two sections of the caste who are, no doubt, not far above the village scavenger, but most of the tribes have adopted subdivisional names which obscure their affinity to the parent stock, because the latter has but a poor reputation amongst the settled population of the plains of Hindustan. Here the D6m is a vagrant, and nominally does cane work for a living. He also makes leaf nlatters and blows horns at marriages, which is a performance never attempted by any but the lowest castes. In Assam he is a fisherman. In the Sewalik and Lower Himalayan valleys, as far west as Kashmer, the D6m, or Dumna, is also an artisan or exerciser, and in the latter capacity again associated with musical instruments. But the highest type of D6m is the Mirdsi, or genealogist, of the lower Brahmanic castes, and, as such, their bard or minstrel, so that his musical gifts, especially with reference to the drum and timbrel, again come to his service. In the South Deccan and northern portions of Madras, the Domba is, generally speaking, an acrobat, but some few of them seem to be weavers of coarse fabrics. There are many links showing the connection in race of the Dom with the Mang, Mahar, or Dhed of further south on the one hand, and on the other, with the various tribes of semi-criminal vagrants of Upper India, such as Sansi, B^dia, BAwari, Kanjdr, Changar, &c., and to the foreign investigator the chief interest in their ramifications lies in the probability of tracing to some of them the parentage of the gipsies of Europe. With the exception of certain classes of the Bawari, the above-mentioned tribes are all vagrant, whereas the Mahar and Dhed are all settled, and of the Mangs only the small section addicted to snake charming and conjuring still take to the road. But the question of the origin of the gipsies of the west appertains rather to this part of the subject than to 201 that of the vagrant group of castes, because attempts' have been made to derive from the title D6m the universal appellation of R6m giveu to their community by the gipsies, and meaning, of course, according to their interpretation, " the man." There is every reason to believe that the gipsies are really thfe descendants of the 12,000 or so of musicians and earthworkers, &c^ transported to Persia, probably from Sindhj ill the time of Behram Gaur, about 400 A.D., and in the section of this chapter dealiiig with languages, it was stated that the title Egyptian was probably derived ffom Jat, the prevailing caste of Sindh and the Lower Pan jab at that time. The reas6n for adopting this root is mainly the insufficiency of evidence for the alternative derivatibn. In all 'the varieties of the gipsy language in Europe we find lots of low Prakritic words, a good deal synonymous with those in use in Upper India iix the present day, but scarcely any Arabic. Again, there is no tradition regarding the passage of the gipsies to Europe by Egypt, whilst there is a very definite one as to their migration from Mesopotamia and Persia to North Syria, and from thence to the^hores of the Bosphorus. • Now, in Syria the gipsies are called Zatti, or, in the plural, zatt, to this day, and Arabic authors derive this term from " Jat, an Indian tribe," and give the same title to a particular kind of cotton cloth woven by the gipsies. There is, again, the statement, though the foundation for it is not given, that in olden times, Bithynia, where the gipsies undoubtedly settled, was called Little Egypt, but it does not seem that this derivation of the word is required, in the light of the evidence available as to the connection between the gipsies and India. Language, customs, caste-exclusiveness and all, point to their origin in the Indus Valley, though, as Browning says of them : " North they go, south' they go, troiiping on lonely ; And still as they travel far and wide, Catch they and keep hold a trace here, a trace there, That puts you in mind of a place here, a place there." But though we admit this much regarding the Indian origin of the gipsies, the alleged connection between D6m and Il6m seems to be doubtful. It is open to question, in the first place, whether the interchange of E, and a palatal D can take place at the beginning of a word, though it obviously can in any other position. Then, again, there is no reason to affiliate the gipsies to the D6ra rather than to the vagrants more specially given to the performances for which the gipsies are cele- brated, who are found in the tracts from which the latter were historically recruited.* From another point of view, that of language irrespective of ethnology, we find that E.6m means, not only man, in the general, but, as in Grerman, married man or husband, in the particular, so we may connect it with the Prakritic Raman, a husband or lover; or the same word, in its earlier signification, of restless wandering ; or, once more, in its sense of sport, amusement, as the change of a vowel costs less than that of a consonant. Mr. Leland's authority ,-|- John Nano, maker of curry powder in London, but by origin a " Mahometan Hindu " of Calcutta, seems to have been akin to the " certain persons of Kyr^ne," whp favoured Herodotus in a somewhat similar manner. A great portion of the trading community returns itself by the title of Bama, Vaishia, or Mahdjan, which are simply functional designations, and ignore the real subdivisions of caste which underlie the above. Thus, the full strength of many of the items shown in Table XVII. (A.) under this group is not that, given in the return, but something considerably above it, the balance being included under the general title. We have, however, a few definite entries. Ardra, for instance, is the great trading and shopkeeping caste of the Western Panjab. The Khatri, its neighbour to the east, is not so exclusively given up to trade, and, like the corresponding class in Gujarath, furnishes a considerable portion of the stafiF of employes in the offices of the local government. The commercial element of Telingana is supplied by the Kdmti aud Balija, and possibly some of the former have Class VIII. Traders, &c. 12,270,973 Group 1. Traders - 12,148,597 1. Mahajan and Bania (itn- 3,186,666 specified). 2. Agarwal - 354,177 3. Khatri . - 686,511 4. Ardra 673,695 , 5. Komti - 545,206 6. Balija 804,307 7. Chetti - - 702,141 8. Mappila 916,436 9. Labbe - 364,293 10. Lohana 530,468 78388. * A brave attempt has been made to derive conjure from Kanjar.- + Gipsies, p. 337. ■ C -J. A. B. 202 been inoluded in the latter, wliicL ]jas a wider applic?ition. In the south,, the inland tracts are served by the C/iei^i, as the Kanarese are by the Banjiga, again a general term, covering numerous subdivisions. ' On the south-east coast of the, peninsula we find trade much in the hands of, the Labbe, a Musalman tribe, of milled, origin^ and, on the Malabar tract, the .Mappih occupy a like position. The Igitter, however, are, in great measure, agriculturists, and having acquired estates have abandoned commerce, but, as a class, they may still be considered to follow the latter* Their disputes with the Nair and , other agricultural classes, due probably, as in Elg-sliern Beng^lr to their origin amongst- the lowest Brahmanic castes,, and their, religious zeal, have brought them into serious trouble on several occasions during.the last few generations ; so much so that special legislation had to be applied to their case, and the " Moplah " outbreaks were the main political feature of the coast. For t^e lastlO years, however; matters have been quiet, though the devotion of: the community to their faith is no less marked than before. , , ; , The Lohdna of Sindh are by no means a well-defined body, and probably many of those shown under this title are not of the commercial subdivisions.. Like the Khatri, they fill the Government offices, as they are almost the only literate " com- munity in the whole province. Amongst the smaller trading castes of the west coast, we find the Bh^tia, located chiefly in Kaohhand Sindh, with,som« brethren, of the same name in the Panjab. They are celebrated for their enterprise in foreign trade, especially that with East Africa, and for their sectarian orthodoxy with regard to the Wallabhacharya denomination, of which they, are - amongst the chief supporters. Then, again, the Meman, a Musalman body, converted from Brahnianism, is .found side by side with the Bhatia, whilst the Kh6ja, who are of the same origin,' but different in their Musalman form of faith, extend well into the Eanjab. It is doubtful, however, whether the Kh6ja of the Peshawar 'yalley and its neighbourhood are of the same race as the merchants of this iianie on 'the Boiiibay coast, though they may have the same reverence for the deified' descendant of the " Old Man of the Mountain " of the time of the Crusades, who has been for ; a couple >bf generation or so domiciled in Bombay. la Grujarath and the North-West Provinces, the subdi^^isions of what is returned as Bania are exceedingly numerous. The best known are the Agm^wdl, from the north, and the Oswal and. Porwal, from Eajputina, with local varieties, such as the Kasarwani, Shrimali, and so on. The Deccan, again, has its own variety of trader, in the Lingaiat ,and its subdivisipns, and the.Konkan coast sends out three or four subdivisions returned under the question-beggiiig epithet of Vaishya. As a rule, however, the more extensive operations in the Maratha country are in the hands of Bania from Gujarath or Marwar. In Bengal there seems a great reluctance to retain local variety, so most appear as Bania or Vaish. The pedlars are a small and scat- tered group.! The only large caste is that of the bead and bangle sellers, Manihdr, and these, too, contain both Brahmanic and Musalman families. Peddling, however, is now carried on by many of the trading class 'and by Musalmaiis who do hot return the caste. Geoup 2. Pedlars 1. Manihar 2. Thoria The hieratic castes, to which our Class IX. Professional 31,652,422 Group 1. PriesU ■ - 15,467,752 1. Brahman 14,821,732 2. Jangam - - , -^ 396,598 3. Ulama ,- - 50,166 4. GarMi " - 41,4l'2 Group 2. Devotees andj^scetics 2,717,861. 1. Gosain 231,612 2. Bairagi 275,604 3. Sadhu (iinspecified) 376,130 4. Bawa „ 66,115 5. Vaish nav 469,052 6. Fakir 830,431 Gkgup 3. Temple Service 32o,53q 1. Gurao n 0,529 2. Satan i 88,354 3. Sewak, &c. 1 121,647 attention is now directed, constitute anything but a homogeneous, group. In the first place, the vast majority of' the population herein included belongs to conventional, not functional, bodies. The other four, it is true, are recruited with reference to the duties they perform for their respective Qommunities. The Jangam, for example, is the sacerdotal class of the Lingaiat sect, and owes his name, it is said, to the practice of the votaries of that sect of wearing the emblem of Shiva, their eponymic deity, in a locket, on their persons, instead of being satisfied with locating it in a temple or shrine. They thus make it "jangam,? moveable, instead of " sthawar," fixed. But their religious guides are not in the present day entirely dependent on their congregation ^oi'.4^eir support, and are wesll Jinpwn as industrious traders in grain and other village prodliQe thi^ou^Eout ' the Sdutli Deccan. T£.p Ulama are the religious teachers °^ ^*^^;^^iiJ^b..Myi8alman; ,pd are.notfotujid returned under this title in other provinces. THeGArudais a, small body, but ; has' been mentioned as repyesentiijg the priesthood of the village menial, who, is prohibited from participatiqi; in the Gpfempnialof the classes above him/ The caste seems ■ to , lie confined to 'Eajputana and;- the, tr^(^'t immediately to the south-west thereof,, Jja'^^H^ there are over ,^0,000 persons returned "as Pandaram by caste, 'which title' probably represents those whoare the hereditary or elected administrators of large temples, or who perform religious services in connection with tjie ' deities in honour of whom the temple was erected. It has been already mentioned that very; early in the Indian portion of the histpry of the Arya, the Brahman ceased, to perform these functions, and the few bhat continued to participate in them were held to be thereby degraded. Some of these sub-castes of Brfihmans are still in existence, but, as a rule,, the working staff of the temple is not Brahman, though, somehow, the, latter, caste is always at hand when largess is in question. _ In the Brahman caste, returned as such, we have every sort and grade of subdivision. In the Panjab there is the Muhial, whose aim is military service, like, the Pand^. of the G-angetic basin in ^Hindustan. . The cultivating classes of the Desai of South Grujarath and the corre^pjonding class in Orissa are both termed Mastan, and this title is shared by other >castes of cultivating Brahmans in Upper India. Some of these last perform the whole cycle of operations connected with tillage, whilst others draw the line at holding the plough, and employ their serfs on that task. There is a congiderable number of Brahmans engaged as family and village priests, but the majority has taken to secular pursuits and been subdivided accordingly. In Hindustan the number of Brdhman cultivators is very large, and on the west coast, both in Malabar and along the Konkan, this caste, in various distinct communities, is prominent amongst the landholders. In the Deccan, again, the Brahman almost monopolises the occupations, barring trade, that require reading and writing. In law and education they are everywhere to the fore. Some have taken to the surveying and outdoor work of civil engineering, but medicine, for ceremonial reasons, has not yet made much way amongst them, and the few that have advanced any way in that profession are scarcely regarded as orthodox. But whatever may be the means that, a Brahman adopts for gaining his livelihood, and how€!ver it may be regarded , by ;his fellow-Brahmans, his attitude towards the rest- of the community remains i^nchanged, and the distance .ibetween them is never allowed to be diminished. The caste is seen to differentiate in all directions within itself, but towards outsiders^ it presents a, honjogeneous and. unbroken front, and claims the privi- leges of birth, semper, ubique, and. ah.OfMhibfis.,, In its capacity of the literary caste of the people, it will be discussed in the next chapter. , . We have, then, the devotees and ascetics, of whom there are many groups, none, probably, represented "m the returns at its full strength, by reason of the attraction of the general title of " holy man," " brother," and so on. Out of the whole group, nearly a third are Musalman, chiefly found in the Panjab and Hindustan, The bond between the fraternity, of whatever subdivision, is the rule not to live except on charity. In the case of the Gosai, however, it is not al-vfayfe the case- that cucidlus facit monachum, and we find that the horsehair plaits and the portentously large rosary belong to a shrewd trader, an industrious cultivator, and, in the pre-British time, to a fanatic and indomptable fighter. But, as a rule, the caste goes from shrine tO' shrine, living on alms, and giving in return the l)lessing to which this course of life seems to impart such efficacy. The temple servant is returned as a caste almost entirely from the Deccan and Madras. In the former, the Gurao has also the monopoly of making the leaf platters used at caste feasts, and so on. The Satani of the south is apparently a cultivator also, and is found in considerable numbers in the returns from Burma, whither he immigrates as one of the numerous seasonal labourers. In Upper India, and Bengal the services performed by these castes are relegated, apparently, to others, who were not specially given up to these occupations at the time the caste was first evolved. The next group comprises the numerically small but interesting and widely- spread castes whoare engaged in retailing by tradition the ballads recounting: ■ the annals of the line of chiefs at whose courts they are entertained. In this way they are the genealogists of such families, and such. is. the veneration of their profession, or, c 2 Geotjp 4. Genealogists : '1. Bh,at - 2. Ohacan ■' 5go,4.i2 481,119 99,09() Gkoup 5. Writers 2,555,867 1. Kayasth 2,239,810 '2. Karan 146,053 3. Karuam 54,177 4. Kaunakan 41,013 5. Piabhu 29,659 6. Vidhur 33,437 204 perhaps, the dread of the loss of the tradition in question, that their persons are as inviolable as the laws of Mann would make that of the Brahman. There are two castes of note in this category. First, the Bhat, or Bhardt, who is sometimes further distinguished by the addition of Eaj, or Royal, bard. It is curious to see that he is not found in such profusion in the heart of Rajasthdn, where reside the descendants of the heroes of many of the most renowned of his ballads,^ as in the North- West Provinces, where the Eajpiit is twice as numerous, and in Bihar, in both of which the blue blood of the Kshatria has suffered grievous dilution. The Charan is in all probability an off-shoot of the Bhat, differentiated when the latter took to genealogy alone, and got above ballad singing, for, in JSTorthern India, at least, the last-named function is appropriated by the Charan. In many parts of the country, too, the Bhat has taken to agriculture and trade, but he is not found far from the castes that most call his traditional functions into exercise. In a former paragraph of this section it was pointed out how the genealogies of the middle classes of Northern India were kept, and their ballaids sung, by the superior divisions of the D6m caste, but neither they or the Charan seem to enjoy the personally sacred character of the Bhat. The writing castes are mainly functional, and have come into prominence with the British system of administration. In old days they were in existence, no doubt, but were kept well in subordination by the Brahmans. For instance, Manu, whose code embodies the aspirations of the Brahman of the revival of that religion, ordains that a Shudra, when other occupations fail, should take to writing for a living, and there are other quotations to the same effect, whilst popular proverbs, too, attest the general distrust of the man who lives by his pen. But a naturally intelligent community have lost no time in taking advantage of their opportunities, and the castes that are not merely the village scribes, as are some of those in the south of India, have not only risen in wealth, but have devoted a good deal of research and ingenuity to proving their right to Kshatria origin. In fact, there seems to be a good deal of truth in the presumption that in Northern and "Western India, at all events, the position of the writer at the courts of native chiefs in old times was due to their left-handed connection with their patron's family. But such a connection has not been acknowledged by the rest, and one of the complaints most loudly and frequently heard from the well bom of the community is that under the present system of administration, — " The bfiggar's book " Outworths the noble's blood," so that the affairs of State are falling into the hands of castes, who, in private life, are not even admitted within the portals of the social leaders of the people. Of the writing castes of this class the most important is that of the Kayasth, which is found chiefly in Bengal and the North-West Provinces. From the former it has migrated into Assam, and in both it is used as synonymous with writer, and is thus recruited from the lower castes, who, as education spreads, take advantage of the term to escape from their origin. In the north-west the caste seems to be more exclusive, but is not altogether devoted to the pen, and furnishes some proportion of the grain parchers and tailors. In the struggle for recognition in the social rank to which it aspires, the Kdyasth caste does not return its full strength at the census, but the Kshatria, under various general titles, such as Surajbansi, and so on, gives shelter to a good many. The Karan, or Mohant, of Orissa, is a caste of the same nature as ' the above. In the west of India we find not only the Kdyasth as a strictly maintained caste, but the Prabhu, an entirely local development, though assigning to itself a Kshatria descent. A still smaller caste goes still further in this direction, and calls itself Brahmakshatria. None of these castes are functional, and as they are well-to-do, they have been able to maintain their position without admixture with writers of other origin. The Yidhur of the North Deccan and Central Plain is, no doubt, a compara- tively recent offshoot irom the village Brahman. The Kannakan, sometimes adding to his caste the quasi-surname of Pille, is the descendant of the Conicopillay of Orme. The Karnam is the caste of village accoujitants in the South and Bast Deccan, which succeeds, m tnose tracts, the Deshasth BrAhman of the Maratha Deccan. 206 The class we have now to deal Class X. Arts and Minor fessions. Pro GrBOUP 1. Astroloffers, Sfc. 1. Jdshi 2. Vaidya 3. Dakaut 4. Kanisan Group 2. Drummers Ballad Singers. 1. Mirasi 2. Dafle, &c. 3. Gondhali and Group 3. Dancers. Singers Group 4. Mimes, Sfc. 1. Bahurupia 2. Bhand 3. Bhaiia with is the minor professional, such as the musician, as he is found in India, and singers and dancers, the two being combined, as stated in the chapter on occupations. The first group, however, in the class is of a different type, and comprises the village astrologer or horoscope caster, with the hail averter, and, in Bengal, the medical caste, which does not seem to be found, as a separate entity, elsewhere. There is little to be said of this group, as the main caste^ are probably functional, and recruited from subdivisions of others. For instance, the Joshi is sometimes a sort of BrAhman, and so, in Upper India, is the Dakaut. The hail averter, Gurpagdri, is chiefly found in Berai- and the Nagpur or Maratha districts of th^ Central Provinces. The next group is a large, and, as mentioned in connection with the D6m, a disreputable one, without being criminal. The Mirasi is found almost entirely in the Panjab, Rajputana, and thfe ISTorth-West Provinces. The Dafle and their instruments of predilection, are the village drummei'iS north and centre. In the south, others, from the Paraiy& on this instrument. The Gondhali are the ballad occupied by the lattejr and 4,153,275 299,776 85,306 87,193 16,062 27,198 645,214 316,422 147,364 18,034 145,778 62,507 4,940 9,783 24,539 Class XI. Carriers 1. Banjara 2. Labana 3. Pendhari Dholi, called so from in most provinces of the and allied castes perform singers of the Marathas, and do not go beyond the tracts .. ^ caste. In the next group, that of singers and dancers, no items are shown, becausle the body is recruited from many castes, properly so called, and the general title usejd largely in the census schedule is simply that of the function, as Tawaif, KaMwant, c^r Kanchan. In like manner, the small class of the mimes and puppet-show keeper's is made up of a number of names which are purely local, with the exception of the twjo mentioned first in the marginal table, which are found in minute communities iri most of the northern provinces. The next class consists, practically, of two castes, and "probably the second of these is but an offshoot of the first. The Banjara, otherwise known as Brinjari and Wanjara, seems to be traceable to the forest grazing lands at the foot of the SewAlik range, and along the base of the Himalaya, in the north of Rohilkhand. Their origin is unknown, but they appear in literature in about the 10th century of the Christian era. Their name is said to be the same as that of the Kanarese castes of traders, Banjiga, from the Sanskrit Banijia-kdra, merchant. All over Hindustan and Eajputdna they carry loads on their bullocks to and from the coast. In the North Deccan they have settled down as agriculturists, though here this caste denies its connection with the wanderers. In the Panjab the caste most prevalent is that of the Labani, which is probably only a subdivision of the Banjara. The same term is used throughout the South Deccan and the central portions of Madras. It is curious that the name should be the same as that of the writing and merchant class in Sindh, and of a caste of shopkeepers in Gujarath, and that the Labana of Central India should assert their Gujardthi origin. The coincidence seems to indicate a , connection with the Gujar Skythians before the latter had been dispersed across the desert. There is a further distinction between the Labdni and Banjara of the north , and the same caste in the south, namely, that the former are either settled residents or well-behaved characters, whilst the latter are of very evil repute, and constantly on the move to avoid intimacy with the local police. The few Pendhd-ri that appear in the returii are all that are found to return themselves by this once celebrated 1)itle. They seem to be carriers by origin, and to have affected the Deccan and Central India. i^ear the latter tract, and partly in it, the marauders of the beginning of the present century founded q. small State, which is still in the occupation of their leader. They are jifusalm^ns by religion, and have no doubt returned themselves chiefly under one of the foreign titles* Cc 3 206 In this next class we have a very varied doUection of castes, some of them reputable and useful members of society, and others just the reverse. As a rule, the latter predominate. Most of the items are small, and on comparing the various titles it appears that many are but linguistic varieties of the same tribe, or that in changing his sky the vagrant has also shifted his occupation. The first group is respectable, in spite of his use of the donkey in carrying his! wares. In a country where grain and pulse form the staple diet, the grindstone and the curry tablet and roller are too important implements not to be well looked after, aiid the two castes named in the margin do not represent the full strength of those who keep these articles in order. The second group includes the stone quar- riers and earth- workers. The Od, or Waddar, is probably the parent caste of both, and where it is prevalent, as in the South Deccan, it is subdivided into those who dig and those who quarry, or split rock a la Hannibal, but using a less costly liquid. The origin of the caste is uncertain, but if not from Od, the old synonym for Orissa, it is likely that Telingana gave it birth, as even in the Panjab this caste preserves its southern habits, and wears the black blanket peculiar to the Deccan. The name of Belddr IS evidently a northern one, and has ex- tended .south with' the migrations of the tribe which use the implement it represents, a peculiarly shaped spade or mattock. The knife grinders are represented by one caste, and that a small one. The occupation- is carried on in towns by subdivisions of the blacksmith castes, and the Shikligdr is the only vagrant, and even he is said to be settling down in fixed abodes. The next two groups are hardly separable one from the other. Mat making is too often a cover for nefarious observation of the habits of the villagers of the place where the caste in question pitches its camp. The fowler, too, has the reputation of being a noted snapper-up of unconsidered trifles. The cane worker, pure and simple, such as the Buriid, is often respectable enough, and leaves his village only for the travelling season, returning during the rains. Many of the resi, travel for purposes distinctly immoral, and are constantly being " moved on " at the request of the resident population. They are, moreover, closely connected with the great tribe of Kanjar, or Sansi, in Upper India, with the Bedia, in the tracts more to the east, and to the Jogi, who, under this philosophical appel- lation, travels all over India, telling fortunes, singing songs not over decent, and swindling the cultivating public by most palpable frauds. Marco Polo says they live to 150 and 200 years, but, fortunately, the receipt for the draught of sulphur and mercury, by which they achieved this longevity, has apparently been lost. The last group comes from exactly the same stock, and many of them have all the aptitude of hereditary training for conjuring and acrobatic feats, in which they are remarkably proficient. It is superfluous to enumerate all the, items, mostly small, that come under this head, but they will he found in Table XVII. (A.). The comparatively large group of castes, the titles of which are either unrecognisable or so indefinite that they could not be classed, concludes the list of Indian social divisions. In the last category, it may as well be explained, for the benefit of the readers who are acquainted with India, are placed such entries as Madras! Bengali, Zamindar, Outcaste, and so on. • Class XII. Vagrants • 3,457,666 Group 1. Grindstone Makers 18,996 1. Takankar 2. Kliumra 9,508 6,554 GrEOBP 2. Earth and Stone IVorhers. 1,124,357 1. Od, or Waddar 2. Beldar 793,516 152,515 Geodp 3. Knife Grinders 18,980 Sliikligar 16,781 Group 4. Mat and Cane Workers. 639, 1 5o 1. Korvi 2. Buriid 3. Bansphor 4. Basor ^ - 5. Changar 6. Ghasia 207,045 53,413 89,955 73,345 36,569 46,077 Group 5. Hunters and Fowlers 948,870 1. Wagri 2. Bahelia 3. Mahtani 4. Mdghia 5. Valaiya 6. Aheria 179,070 39,203 56,984 146,667 289,411 36,320 Group 6. Miscellaneous Va- grants. 400,969 1. Jdgi 2. Sansia 3. Kanjar 4. Barwala - . - 214,546 30,704 29,486 63,856 Group 7. Jngglcrs and Acro- bats. 306,344 1. Bedia 2. Nat 1 65,194 139,068 Class XIII. Titles. Indefinite Indian 3,079,304 207 Class XV. Musulmans of Poreign Title- 34,348,085 1. Shaikh ,- 27,644,993 2. Pathan 3.225,521 3. Moghal 333,114 4. Saiad 1,430,329 5. Baldch - 971,835 6. Tiii-k 50,503 7. Arab 39,338 Tlie rest Qf the table needs but little comment. The 14th class, showing the Indian Christians,' contains, as explained in the end of the preceding section, only a portion of that community, and the Groanese have only been distinguished in Bombay and a few other. places. The class of Musulman that returns a foreign title includes nearly 12 per cent, of the population, so that it comes next to the Agricultural group in numerical strength. In the mar- ginal statement the chief items, are shown. Out of the 27^ millions of Shaikhs nearly 21 millions are returned in Bengal, the reason for which concentration has been already given. The Pathan, as he is called east of the Indus, comes next in order, and, probably, most of the 970,000 returned in the Panjdb are really of this race, or nearly akin to it. The rest, with the exception of a few colonies in Bohilkhand, have little claim to transfrontier descent. Like the Pathan, the Baldch haunts only the frontier provinces of Sindh and the Panjab. There are, however, a few bodies of Makrani, with a good tinge of African in them, in many of the native courts of Western and Central India. The Baiad, or descendant of the adherents of the prophet, is found in all provinces and States. In the West Panjab the title is given to any holy man of that faith, and means simply a religious teacher, supported by the alms of the believer. Elsewhere the title is assumed on conversion, or, according to the proverb, on the acquisition of the proper amount of worldly goods. In the case of the Moghal, the strain of foreign blood is more traceable than in that of the Shaikh or Saiad in the north of India, and amongst the mercantile classes of the west, where it often means anyone of Persian descent. The Turk is found chiefly in the north-west corner of the Panjab, where several tribes of the race are returned. In Bombay and Haidrabad there are said to be a few members of the Othmanli community, and in the Rampur State, in Rohilkhand, there seems to be a considerable colony of this title. Here, however, it is not certain that, as in Madras and formerly in Burma, the name Turk is not given to any Musalman of foreign origin. The Arabs are returned mostly in Aden and Bombay. In the former they come from the mainland, and in the latter from the Persian Gulf, either with horses from the northern parts of Arabia, or from Maskat. Along the Malabar coast, too, there are settlements of traders of this race, which have been in existence for many centuries. The Arabs in the Panjab are either merchants from Bombay, who have entered into business at Multan, or at Peshawar, where India meets Central Asia, or else are local Shaikhs, who have adopted their ancestry more thoroughly than the rest. The great tribes of the Himalayan Mongoloid race are but poorly represented in the census returns. In the west we have the Ladakhi of Kashmer, a title which includes, probably, the Bdlti, of Skardu, who is found in Simla for a great part of the year. The Bhoti, which is the general name given to those of Thibetan origin, comprise both the people from the British Himalayan districts beyond the first snowy range, as well as the inhabitants of Bhotan, who come in small numbers to Darjiling and the Assam markets. Between these two come the Nipali tribes, of whom we have only the settlers in the submontane tracts and the Tar^i grazing lands and forests, together with the comparatively few soldiers in so-called Grurkha regiments, many of whom return their caste as Kshatria. There are also immigrants for labour to be found in the tract round Darjiling. Class XVI. loids. Himalayan Mongo- 1. Ladakhi 2. Bhoti 3. Lepcha 4. Khambu 5. Tharu 6. Xewar 7. Lirabu 8. Giiriing Mangar Sunuwar Murmi 9. 10. 11. Class XIV. Indian Christians 1. Native 2. Goanese Class XVII, mese. Burmese Karen Shan Chinese Arakanese Yau M(5n Ah 6m Cross breeds Assamese and Bur- 1. 2. 3. 4. 6. 7. 8. 9. 244,722 30,672 25,670 9,745 33,490 53,875 4,979 15,079 10,894 19,383 5,210 21,889 1,836,848 1,807,092 28,756 7,297,618 5,408,984 540,876 182,745 41,832 452,164 12,934 467,885 153,518 19,821 The Lepcha are mostly from Oc 4 208 Sikkim, now resident in Ddrjiling. Of the Tarai settlers, the KMmbu and Limbu admit their Kirant or Himalayan origin, whilst the Tharu apparently repudiate it, and it is possible that their Mongoloidic type of feature is due to cross-breeding with the Mech and other forest tribes, after they were expelled, as they assert themselves to have been, from their realm in the Gangetic basin. The I^ewar are the people of the Nipal Yalley displaced by the Giirkha, and the Guriing, Siiniiwar, and Mangar belong to the military tribes of that State. The Miirmi, like the Newdr, seem to have been once powerful in the higher valleys, and to have succumbed to the rule of the new- comers from the plains. Of the rest of the classes sufficient has been said above, in connection with language and religion, so with this may be concluded a very lengthy, albeit inadequate, description of the chief elements in the heterogeneous mass that is known as the " people of India." Class XVTTT. Western Asiatics — 1. Jews 2. Armenians 3. Parsis 16,951 1,295 89,618 Class XIX. Eurasians 81,044 Class XX. Europeans 166,428 Class XXI. Africans 18,776 209 CHAPTER VI. The Distribution of the Population by Literacy. , 1 The wavea of the ocean, before reaching the child, break against four walls, which encompass the water of his education or crystallisation— father, mother, brothers and sisters, and a few extra people, are his forming- world and mould. But, all this deducted, we must remember in education that its power, like that of the spirit of the age, which must not be measured by individuals but by the concentrated mass or maijority, must be judged, not by the present, but by the intuve.—Richter. Where the task of public instruction is undertaken by the State to the extent that it is in India, the function of a census of Literacy is to supplement the current record of progress in regard to this important matter. This can be done either by taking stock of the results on the given date, so they may be shown cumulatively, instead of merely by annual instalments, or by bringing to book the out-turn of institutions which are territorially or departmentally beyond the scope of the annual reviews. How far the census fulfils this object will be discussed in the course of this chapter. The first point to be considered is the scope of the inquiry made at the census. With respect to the tracts brought under enumeration, it was not thought worth while to attempt this detail in the wild country on the eastern frontier of Bengal and Assam, or in the very similar tract to the south-west of Orissa. Then, again, in Rajputana and Central India, the establishments available had to be treated, as explained above in connection with the return of mother tongue, with the utmost leniency, so the schedule was relieved of this column also. Lastly, in Kashmer, as the census was being taken systematically for the first time, it was held that more important statistics than those of instruction should alone be, recommended. We have thus to deal with a population of 261,838,926, detailed in Table IX. of the Imperial series. In 1881, nearly the same omissions as on this occasion were allowed in the Hill tracts of Eastern Bengal and Orissa. The Malabar States and those under the Central Provinces, as well as Khairpur under Sindh, and the Naga and Garo Hills in Assam, were also excluded, but have been brought on to the record of the present census, in addition to Quettah, Aden, and the Andaman settlement and the newly acquired territory of Upper Burma. To show the nature of the information asked for, the rules for tilling up this column are here reproduced : — 1881. 1891. " (joI. 12. Against those under instruction write i " Col. 12. Enter in this column against each per- 'is learnino-'; against those not under instruction j son, whether grown up, child or infant, either but able to' read°and write, enter 'knows'; against j 'Learning,' 'Literate,' or 'Illiterate.' Enter all those who cannot either read or write, or who can I those as Learning who are under instruction, either read but cannot write, or can sign their names but ! at home or at school or college. Enter as Literate cannot read write ' does not know.' Only those j those who are able to both read and write, but are should be shown as able to read and write who can not under instruction as above. Enter as Illiterate do both." I those who are not under instruction, and who do not ' know how to both read and write, or who can read but not write, or can sign their name but not read." " Col. 13. Enter here the language which those I shown as literate in col. 12 can both read and write, and if a person knows how to read and write English as well as a vernacular, enter English also." " This column is to be left blank for those shown in col. 12, as Learning or Illiterate, and, except when English is known, only one langdag* should be entered, namely, that best known." Thus three classes were distinguished. First, those who were under instruction ; secondly, those who had finished their schooling; and lastly, those who had not had any Special provision was made to exclude from the category of the pupils those who attend what are known as " Rote Schools," which are numerous m the Panjab, and are found nearly all over India to some extent. In such establishments reading and writing are not taught, and the curriculum is confined to the inculcation of portions of the Kurdn, with the appropriate enunciation and gestures or of some Purlnic lore both entirely by oral tradition. Then, again, m the present day so many messengers porters, and other menials find it to their advantage to be able to sign S names, that they acquire this amount of literature without ever ^advancing I 78388. D d 210 bejond it; and it was held advisable to specially exclude tbis class from the category of literate. So far, tbe instructions in 1891 followed those of the preceding census. But it was thought worth while to attempt a further step by adding to the return of the literate, supplementary' intolfTcahtion regarding the number that were conversant with the. English, language. This inforination, it will be observed, was not called for in respect to those still under instruction, whether at school or college, as the object was to include only such as were, so to speak, out in the world. Taking the instructions as a whole, experience in every Province has shown that the distinction between those under instruction and those able to read and write, but no longer in a state of pupilage, is one which it is advisable to abandon at a future enumeration, since there is a general .tendency to disregard it. The former class is considered ijlferior to the latter, and as India is, according to Jacquemont, " I'Utopie ,de I'ordre " social a I'lisage des gens comme il faut," the dignity of the castes to whom literacy, according to that social ordinance, should be restricted, revolted at the notion that the half-;naked iirchin set tq drive his father's cattle as soon as he has completed the fourth vernacular st9,ndard should be entered as Literate, whilst the Brahman reading for his degree in Arts was relegated to the category of Pupils. The feeling is wpU expressed by the Census Superintendent for Mysore, himself a Brdhman and experienced in educational matters, who writes : — "Moreover, the enumerated persons returned themselves in a vast number of cases as learned, though they were stjlL^anM'w^, probably in. a spirit, of bravado. For it is a notorious characteristic of the student of tjie period that he is generally too obtrusively and superciliously self-conscious of the superior educational advantages enjoyed by himself, and as his contemporaries and seniors of the old school were returned as literate, his persobal vanity was not unnaturally tickled, and many of his class returned ' themsel'*fes at the enumeration as fear*je«?, although still fearrafo^, thereby exaggerating the ranks of the literate, at the expense of the numbers upfler tuition." The results of the enumeration were tabulated by the three age-periods, (a) under 15, (&) 15 to 25, and (c) 25 and over ; so it was not difficult to see that Mr. V. N. Narsinghaiangar is right in his facts, whilst his professional knowledge of the class in question allows the presumption tliat he is not wrong as to the motives he attributes. It is not only in Mysore that the egg, to use a German proverb, holds itself wiser than the hen, and not alone in the Eome of Persius. Ingenium et rerum prudentia velox Ante pilos venit. Thus the really trustworthy division of the population in respect to literacy, so far as the census return is concerned, is into those who_ do not know how. to read and write and those who Table A.— General Summary. do, whether the latter be still under fnstruction pr have ceased to be so. But fbr the sake of comparison in a later part of this chapter with the annual statistics published by the State Educational autho- rities, the three groups* as collected from the schedules, , are shown in the marginal abstract of , t)ie general return now under review. But the depth to which literacy has filtered through the community can be better appreciated when the figures are -set fol-th' proportionally, especially if the above-mentioned age-periods are also recognised. This is done accordingly in the following statement : — Table B. — Proportional Abstract. Total. Males. Females. Learning Literate' Illiterate Total 3,195,220 12,097,530 246,546,176 261,838,926 2,997,558 11,554,035 118,819,408 133,371,001 1 197,662 543,495 127,726,768 128,467,925 1 - J.i J ' J '- -- ' ' : Total. ^ . Under 1.5.. ; , 1.5, to 25., 25 and over. Both Sexes. Males. Females. i^lemales to" 1,000 Males: Males. Females'. Males. Females. Males. Females Learning Literate ^ - . Illiterate 1-22 4;fi2 . 94-16' .'.Il . . 3-25 8-66 89; 09' 0-15 jr. 0-42 99-43 63 47 1,07.5 ' 4-74 l-'27 93-99 e-36 0-18 99 -46 1-96 13-10 84-94 6-06 /i .' ; - 0-85 99-09 . 0-09 13-73 86-18 6-01 0-47 99^52 ^11 Males. 1. Females. ' Learning. Literate, Learning. Literate. Under 15 15 to 25 25 and over - 84-01 14-27 1-62 5-84 24-76 69-36 89-82 .6^89 3-29 , ,16-15 34-55 50-70 Now, the salient facts illustrated by these relative numbei's ave, first, that 'in the whole .population dealt with only 58 persons in every thousand can read and "write, or are learning to do so, and, secondly, that of those 58, 53 are males and five #e of the other sex. If we consider, first, the males only, it will be seen that of those uiider 15 years of age, 94 in every 100 are neither learning or literate, but if "we omit from the calculation all boys' under five, assuming them to be below /the age' at which instriiction may -be held to begin, the ratio of the illiterate falls to 90 per cent.'; After the age of 15 instruction begins to have an appreciable weight, and betweent that fage and. 25, 15 percent, instead of 10, are, to put it negatively, not illiterate. : It is in this period that the confusion between pupils and literate is perceptible, and the two classes are therefore considered together. The next division of the table treats of men of full age, and the proportion of the illiterate rises a trifle. The cotnpara- tively large number of pupils in this period may be attributable in part to the technical and training institutions, or, at least in Upper India, to the adults attending religious lectures in connection with masjid (mosques) in the case of Musalmans, and charitable institutions of a corresponding character fol' Sanskrit studies in the case of Brahmans. The technical name for the former, which would no doubt be returned at the census, is " Seekers' after Knowledge " (Talib ul 'ilm), while the latter would in like manner be termed " Disciples." Another light in which to view these figures is that thrown by the relative ■distribution of ^.the learners and literate by age, instead of that of the age of literacy, which we hav« been considering above. IT ABLE C— Instruction by Age. The marginal table gives this informa- tion. Of the male pupils, 84 per cent, are under and 16 over the age of 15 years. The literate of 25 years and over bear the proportion to the total littrate of 69^ , per cent., which indi- cates, as it were, the cumulative, results of the system of instruction,: Were it not for the intrusion of the literate- , pupil element above, noted ; this ratio would no doubt reach 72 or 73 per cent. The figures for females will be dispussed below. What has preceded this chapter in the present review has no doubt prepared the reader for the great prevalence of illiteracy indicated by the above statements. First 'of all we have the occupational bias of the mass of the population, which is set steadily away from literature. The agricultimst in India is in this respect probably not much below his compeers in far more civilised countries, only his numbers are more preponderant in the community. However well his aflfairs may prosper, it is lonff before the landowner gives over personal participation m all the operations by which he has thriven, and after a long ' course of early rising and late return from the field the " swinkt .hedger " is seldom in the psychological conditions for the assimilation of more abstruse knowledge than is current in the friendly circle that everv evening squats down to discuss the hiikka and pan of village sociabihty. Interested as he is in every novelty that his eye has learnt to perceive, the average raiat is not given to going a step out of his way to discover, any ; but when the three R's are brought within his reach, he is often quick enough to appreciate their advan- tages in dealing with the grain dealer and money lender, so he sencjs his son to school accordingly. Unfortunately, however, as the census and departmental returns show-and experience too often confirms them-the necessities of the farm generally lead to the lad's removal at a rudimentary stage m the curriculum, and by the time he is 18 years old he has forgotten most of the little he learnt when he was 11 or 12. In the towns the state of aflfairs is a little better, but then, as we have seen, their influence extends to less than a tenth of the population. The second influence antagonistic to a more general spread of literacy is the Inup- 'continued existence of a hereditary class whose object it has been to maintain thefr own monopoly of all book-learning as the chief buttress of their social supremacy. Sacerdotalism knows that it can reign over none but an ignorant populace. Tbe nnnn«ition of the Brdhman to the rise of the writer castes has been already mentioned, «nd the repugnance of both, in the present day, to the diffusion of learning amongst the^ masses^ can only be appreciated after long experience. It is true that the Dd 2 Literacy of males. 212 recognition by tlie British. Government of the virtue and necessity of primary education has met with some response on the ipart of the literate castes, but it is chiefly in the direction of academic utterances, which, cannot, in the circumstances, he well avoided. It is welcomed, too, in its capacity of affording the means, of livelihood to many of these castes, as they have to be engaged as teachers, and are bound accordingly to work up to the State standard of efficient tuition. The real interest of tbe castes in question is centred ! on secondary education, of which they almost exclusively are in a position to reap the advantage. This, however, is a topic that will recur in connection with the distribution of literacy by caste, which forms the subject of a later section of this chapter. "We can thus see that the field in which the seeds of literacy have to be sown consists of a few square yards of what we may term relatively good soil, prepared to receive all the seed it can get, and thirsting for the whole of the attention qf the husbandman. Then comes the vast stony waste of labour and menial offi.ces, without sufficient depth, of soil to allow the seed to strike root, and lastly, the many miles of arable mark, so taken up with the production of the food and clothing of the whole community, that whatever else is sown in it is inevitably choked before it can ripen. Literacy of Before taking up the next section of our subject, a glance has to be given at the females. return of the literate amongst the weaker sex. The outlook in this direction is, if possible, more dreary than in that from which we have just turned. It is by the position of women in it that we judge to a great extent of the relative progress made by a community towards the complete life that we call civilisation ; as the Comte de Segur has it, " leur sort est un boussole sur pour le premier regard d'un Stranger qui arrive dans un pays inconnu." Though, as Burma may be said to show us, there may be no intimate or direct connection between the status or influence of woman and her acquaintance with the three R's, a very great divergence, in this last respect, between the two halves of the community more often than not implies the relegation of the wife and mother to the position either of the queen termite, immured for life in the most remote chamber of the ant hill to fulfil her function of replenishing the numbers of the community, or to that of the household drudge enslaved to the caprice of her mother-in-law. It has been said, and we see the truth of the adage not only in India, but in the west also, that she who rocks the cradle rules the empire, and of what sort must the influence be of those, however shrewd and staunch to the better customs in which they have themselves been brought up, who are unable to stray beyond their immediate and narrow surroundings, or get into touch with anything that may help to elevate them on to a higher level ? Their utmost aim is to rear the Telemachus, of whom his father spoke in terms so affectionately depreciatory : — Most blameless is he, centred in the sphere Of common duties, decent not to fail In offices of tenderness, and pay Meet adoration to my household gods When I am gone. According to the current theory, the first duty of a woman is to produce and rear a son to perform the funeral obsequies of her husband ; and, second, to see that satisfactory meals are ready for her husband, and the third, to keep her husband's family gods in good order for his worship. Into such an ideal, intellect can scarcely be said to enter, but faith pervades the whole. On the other hand, it is to intellect alone that an appeal is made by the State system of education, under which come more than 80 per cent, of the whole body of learners, and amongst nations, writes the author quoted at the beginning of this chapter, the head has at all times preceded the heart. But how if the head of the man persists in advancing, whilst that of the women and the heart of both refuses to stir 1 The head is a notorious sapper of creeds based as are those of India, and the " dual life" in that country, of which a good deal has been written of late, is not a feature that one would desire to see extended further than its present sphere, where, amongst the fraction of the 6 per cent, of writers and literary Brdhmans whom it affects, it is comparatively harmless. The social reformer possesses, as Carlyle has it, a prehensile tail. He returns home from his meeting or lecture in full panoply of occidental ethics and political catchwords " up to date," to cast off all these alien swaddling-bands as easily as he divests his shapely foot of its patent leather covering, and his person generally of other encumbrances that interfere with his free enjoyment of home comfort. He performs, ' 213 moreover, all the ceremonies at births and deaths required by the custom of his caste, keeps his brother's widow in the orthodox servitude, and marries off his daughter at eight years old, provided he has no bridegroom ready for her at seven. "Once he has come to forty year," if by mishap he be a widower, he accepts from his friend and co- reformera bride of as nearly as possible the same age as his daughter, and thus tradition is maintained. In these circumstances, there is practically no breach in domestic life caused by the intellectual advance of the master of the house. His studies aro regarded as no more than the proper equipment in life of the class that at the festival of the Diw^li pays homage to its pen as the family bread-winner, just as the trader does to his ledger, and they affect his relations with the outer world only. His indoor life is still regulated by the customs to which he duly conforms, like the philosopher of imperial Eome, who " approached with the same inward contempt, and the same external reverence, the altars of the Lybian, the Olympian, or the Oapitoiine Jupiter." There is thus no want in that life which the education of his ■ wife is needed to supply. Still less is illiteracy felt in the houses of the mass of the population, where the education of the men themselves is, at the best, but surface deep. Against the above remarks there will be urged, of course, the case of the Parsis, amongst the women of which race literacy is far more prevalent than in other Indian communities. The record of the State Educational Department, too may be brought up to testify against the return of the literate and pupils of the fair sex at the census, which is less by nearly one half than the former. This last argument will be examined in a later section of this chapter. As to the Parsis, they form a small homogeneous community of foreigners, who owe everything but bare existence to another race, equally alien to the country, any part of whose customs to which they may be attracted can be assimilated by them unhindered by caste or tradition. There is obviously no analogy between this case and that of a vast and heterogeneous population that has grown up within the country itseU, and has by gradual and histarical process, not by imitation or foreign impulse, hemmed itself in by centuries of the most exclusive prejudices and inviolable custom the world has ever seen. Referring again to Table B., it will be seen that the females imbued with some tincture of literacy come to about six in the thousand, or, as we are dealing with such small fractions, it will be more correct to put the number at 57 in 10,000. Under the age of 15, they come to 54, and the pupils are nearly double the number of those who have ceased their schooling. If we deduct the girls under five years old, the ratio rises to 93, approaching one in a hundred. In the: next period, between 15 and 25 years old, the proportion is nearly the same, or 91 per 10,000, and the literate exceed the pupils. This follows of course, the fact that at this age, amongst the classes that mainly contribute to the returns, a girl, sane and sound, who is not married by the age of 15 is an exception, and would certainly not be paraded at school, even if returned at the census at all. The actual pupils in this period are probably, to a great extent, in Normal and Medical establishments, or widows, and the return shows, too, that a good many are returned from the Christian community, where the age of marriage is later. Lastly, amongst women of 25 years old and over, the proportion of the not illiterate drops to 47 per 10,000, leaving oyer 99^ per cent, illiterate. This seems to indicate that the stimulus to female education was imparted little earlier than the beginning of the present generation. Table C. confirms this view, as not many more than half the literate females come within this age-group. It may be noticed that the proportion of pupils in this group is higher than amongst the males. This may be partly explained by the preference shown by females, in returning their age, for the year' 25 over any year between 20 and 24. Then, again, there are the special and technical institutions above mentioned, and the Zandna teaching, all of which are, probably, at work in Bengal and the Malabar States, where the bulk of the class in question is to be found. A few words more about the proportions of the two sexes amongst the literate and pupils will not be irrelevant to what has just been said. In the population with which tills return deals, there are 963 females to every 1,000 of the other sex. To put it in' another way, in a thousand of the whole population, there will be, on the average, 509 males and 491 females. The females under instruction number 63 to 1000 males in the same stage, but to a thousand literate males. thei:e are but 47 females similarly endowed. The illiterate are more on an equality, for we find only 1 075 females to the 1,000 males, a ratio that would represent an excess of tne former sex of about 116, if the total number of the two were the same. Dd 3 21-t The marffiual Statement D. 2-ives the figurea for 1881 compared with those for 1801 in the Table D. — Comparison of 1891 and 1881. Comparison of Census Return with that of the State De- partment of Public In- struction. m tne same tracts. There will be noticed a slight- dif- ference between 'the latter as here propor- tionately shown and those in Table B., be- cause from the former have been omitted the returns for Upper Burma and the Mala- bar States, where literacy is more pre- valent than it! most of the tracts dealt with, but which did not come within the scope of the census of 10 years ago. As the table stands, -the rate of increase • of the not-illiterate is considerably above that of the population at large, and is especially marked in the case of females. But in relation to the whole population of each year, the leeway that has to be made up is enormous. Where there were four females in a thousand who are not illiterate, there are now five ; and in a thousand males, in place of 91, we now have 104. In a later section,- the variation in the diflFerent Provinces and States will be shown. The next point to be treated of is the comparison of the census return with that published for 1890-91 by the various Directors of Public Instruction in India. Ttis relates to British territory only, except with regard tothe small States under, Pojmbay which appear to be included with the rest. Table B. — Comparison of Census with Departmental Returns of Pupils, Population Distribution per Cent. Per- centage of Varia- tion. 1881. 1891. 1881. 1891. -2 ' ft* 11 ""Learning and Literate. Illiterate L Total 'Learning and Literate. Illiterate Total ' Illiterate Total 10,526,283 105,838,357 116,364,640 432,475 111,332,927 111,765,402 217,171,284 228,130,042 13,416,398 115,454,733 128,881, i3i o89,256 122,735,279 123,324,535 238,190,012 252,205,666 9-05 90-95 0-39 99-61 95-20 10-42 89-58 0-48 99-52 94-44 27-45 9-08 10-75 36-25 10-24 10-34 9-67 10-55 Males. Fern Departmental. ales. Difference of Census. Departmental. Census., Censuii. Males. Fenjales.: Madras 556,449 576,079 87,715 59;i27 -\-. 19,630 ' :- ' 28,588 Bombay and States 551,216 439,360 69,282 30,745 - 1-11,856 — 38,537 Bengal 1,380,385 883,990 88,558 34,845 - 496,395 - 53,513 N.-W. Provinces and Oudh 275,651 238,440 13,870 8,404 - 37,211 - 5,466 Panjab 223,056 158,849 22,657 7,834 - 64,207 - 14,823 Central Provinces 105,699 76,306 5,799 3,901 - 29,393 - 1,898 Upper Burma 49,341 99,229 3,811 3,372 -f 49,888 - 439 Lower Burma 100,668 128,269 14,629 14,853 -1- 27,601 -1- 224 Assam 74,086 49,111 4,698 3,427 - 24,975 - 1,271 Cool-g 4,059 4,192 736 610 -1- 133 - 126 Berar 48,320 38,502 2,022 976 - 9,818 - 1,046 Total 3,368,930 2,692,327 313,777 168,094 - 676,603 - 146,683 The main feature to which attention is required at present is the fact that the departmental return of pupils exceeds that of the census by 25-13 per cent, in the case of males, and as already remarked above, by about 87 per cent, in the case of females. It is further noticeable that the above feature presen'ts itself in every Province, with the exception of Burma and the male pupil element in Madras and Ooorg, exceptions which admit of special explanations. The departmental statement is based on the returns from schools and colleges . in connection with the State, either directly or indirectly, as by grants-in-aid, or by being under inspection, or financial control. The extent. to which non-official schools, especially those of the lower grades, are thus connected, varies in every Province. 215 In the Plan jab, for instanoe, t}ie return includes far more than m most other parts of the Qountry., In- Upper Burma, on the other hand, the number of monasteries not brought on the (departmental books is large, and even in the lower division of the Province, there- appear to be a good many that are equally ignored. As a rule, every Burmese .has to> pass a certain period of his boyhood as a " kyaungtha," or attendant at one of these institutions, even if he may not advance to the noviciate, so the number must be enough to place this mode of instruction within the reach of every village. In Madras, the excess shown at the census is attributed in great measure to the same cause, as it occurs in the tracts such as Tanjore and Malabar, in the firsi. of which there are known to be many elementary schools kept by Brahmans, which are not brought within the sweep of the departmental net, whilst on the Malabar coast, the religious antagonism between the Mappilah community and the rest stimulates the maintenance of many institutions within the precincts of the mosque, wliich are likewise not brought to the notice of the infidel inspector, whether Christian or Brahmanic. On the other hand, at the opposite extremity of this Presidency, the departmental figures for the Granjam district show a number of pupils more than loo per cent, in excess of that found in the census schedules, and this brings us to the second distinction between the scope of the two returns. The census, it is true, includes the pupils of the institutions that do not contribute towards the departmental ■fcotal, but it excludes, by the definition of pupil adopted in the instructions quoted above, all children who are not learning to read and write. The case of the Kuran- Puran schools has been already mentioned, and in the Panj^b we find 57,397 pupils in' them, including over 10,000 of the 22,657 girls departmentally returned. In Bengal, again, this class of institution accounts for over 62,000, and in the whole of the tract covered by the return under discussion, 154,500 come under this category. As few, if any, of .these pupils learn how to read or write, they were, of course, ignored in the census. We then come to the departmental return of " public schools," where 809,116 children are shown as " not reading printed books." Finally, there are the " private schools," with about 262,000 pupils in the elementary stage, out of whom probably the same proportion, at least, as in the other class of institution, are hot yet advanced enough to study printed books. All these, together with the pupils who patronise the rote schools, give a total of about 1,051,000 children in the most elementary stage of instruction, so far as literacy is in question, and as the depart- mental' total exceeds that of the census by 822,286 only, there remains a balance of 228,770 in favour of the latter. In all probability, the excess is far greater, for in Bengal, where the difference- in favour of the departmental return amounts to nearly half a 'million of ,boys and over 53,000 of girls, this return is obviously unduly exaggerated by fraudulent entries made by the hedge-schoolmaster who is a more prominent feature in that province than elsewhere. The same feature appears in the return for the Granjam district of Madras, where the departmental excess, amongst a population notoriously averse from instruction, is most remarkable, and places this tract, in point of education, immediately behind the three foremost districts of the Presidency, Madras, Tanjore, and Malabar. The proximity of Orissa, about which the Educational Inspector writes that "for one case of detection there may be a dozen that go undetected;'' has probably stimulated the local Pandit to go abroad to some purpose. In the Shahabad district of Bengal, the district board practically abolished the system of annual rewards in the year under review, because they "encouraged fraudulent practices." In the Dacca district, in 1890, there was discovered "wholesale fabrication of returns in two municipalities." Even under the very shadow of the central authority, in the suburbs of Calcutta, we find it stated that " it is not unusual for unscrupulous men to start schools at the close of the year for the " purpose of earning rewards, which schools melt away as soon as that object is " fulfilled." In one case, the schoolmaster got the reward by showing off a school attended mostly by pupils of his colleagues, lent for the occasion. . But the prize for ingenuity is taken by the Chittagong master, who held a school m a village on the frontier of the district, so that, having obtained the reward at the annual exammation, he promptly transferred himself with all his pupils to the adjacent village on the other side of the boundary, and reaped his recompense at a second exammation held in his new domicile. History stops at this point, but it would be interesting to learn whether the pupils who shared the toil were also admitted to the fruits thereof. It is irrelevant to the present purpose to inquire with whom lies the responsibility for this state of affairs, it is enough that it exists, and that while it exists there is no reason Dd 4 216 to assume that the difference between the census returns and those of the State department betokens any defect, in the former. Bub apart from intentional fraud, there is the tendency for every village school to attach to itself a sort of kindergarten, ■where the infants of the lower middle classes disport themselves whilst their elder brothers or sisters, who officiate as caretakers, are possibly really under instruction. In the upper grades of this undisciplined contingent, a plantain leaf, a small board with a supply of sand, or even the mud floor, serves as the medium of communicating the alphabet or the digits, whereupon the pupil is duly entered as " under standard." In the case of females this is a very common practice, and in the Panjdb, for instance, where not only a,re half the pupils of this sex under instruction in Kuran schools, where they only learn by rote, but in other schools, more than half are under seven years old, it is clear that there is little to complain of in the small show made at the census. Some of that little, too, is explicable, the Superintendent thinks, by the refusal of rural enumerators to recognise instruction in any other character than the Persian, so that Sikh girls learning Gurmiikhi do not appear in the record. Finally, there is always the confusion between literate and learning to fall back upon, but so far as the comparison with departmental figures is concerned, it is not necessary, on the facts shown above, to have recourse to its discussion in detail. In the provincial reports for Bengal, Bombay, and the Panjab, Messrs. O'Donnell, Drew, and Maclagan have dealt with it in relation to their respective charges. Literacy by religion. Illiterate per 1,000 of each sex. Keligion. Males. Females. Bi'.ibmanic 895 996 Sikl. 904 996 Jain 466 986 Buddhist - 526 974 Parsi 223 499 Jew - - - - 481 786 Musalman . - . 929 997 Christian 657 864 Animistic ... 992 1,000 Total 891 994 The general Table IX. is subdivided according to the main religions, and with reference to a few of the items, perhaps, the subject merits a little comment, though the relative spread of literacy in the community can be better appreciated from the caste distribution, the materials for which are to be found in Tables XVIII. and XIX. in Volume II. of the Imperial series. But the proportion of the illiterate amongst the adherent.s of the main forms of creed will be found in the marginal statement. From this it will be seen that the Parsis hold the first place, both as to females and to males, but, as remarked above, they are only a small and isolated com- munity. With respect to literate males, the Jains come next, as might be ex- pected from their addiction to commerce. But their liberality has not yet extended far in the direction of the instruction of their women, who come fifth on the list, with 14 per 1,000 not illiterate. The second place amongst the females belongs to the Jewess, also the member of a small body, whose male relative comes but third in the roll of his sex. The Buddhist occupies the fourth place, both for males and for the other sex, but in a thousand of the former he shows 474 not illiterate, whereas, owing to the exclusion of women from the Kyaung school in Burma, the corresponding proportion for the sex is but 26. The Christian occupies the fifth place amongst the males, and the third amongst the females. In the former case the not illiterates number 343, and in the latter 136. The four last religions occupy the same order in both sexes. The Brahmanic community comes sixth, with 105 males and 4 females, distinguished from the crowd, but, as was shown in the preceding chapter, the collection of castes that goes by this name represents no uniform feature of irnportance. ]!t is not very different with the Musalman, between whom and the Brdhman comes the Sikh. The latter, in the Panjab, comes between the two in the matter of education only if taken in the lump. Taking separately the larger castes that furnish the Sikh community, these last are more literate than the Brahmanic members, and the balance is restored by the literates amongst the priestly and mercantile classes of the latter. Finally, we have the forest tribes, of whom, as has been explained, only the wilder portions returned their tribal religion. To this section, therefore, literacy can hardly be expected to have extended, so it is not surprising to find but eight in the thousand of the males able to read or write, though it may be unexpected ■ to see that the females are not far from being one in the same category. 217 Table F. But fo^'^c^^. The next topic in connection, with literacy that requires notice is the variety Literacy by in its territoral diffusion in India. ~ in anticipation of this, it may be inter- esting to show how that country stands in this respect in comparison with others of which the returns are available. The marginal statement, accordingly, gives the average number of persons who ai3 not illiterate per thousand of the general population. From this it will be seen that the British colonies in the south head the list, along with the white population of the United States. Less than half the population are literate in Austria, Hungary, Italy, Portugal, Chile, Ceylon, and amongst the coloured residents of the United States. Unfor- tunately, census returns for the German Empire, for Scandinavia, and Great Britain, are not available. The return as far as women are concerned, assumes a slightly different ord^r. The greatest divergence is seen in the case of Ceylon, which illustrates the breach between the West and the Bast in this respect. Portugal, Queensland, and the coloured Country. Number able to read and -write per 1,000 of each Sex. Males. Females. 1. Victoria - - . 2. New Zealand - 3. United States* {White) 4. South Australia 5. New South Wales 6. Queensland 7. Ireland 8. Austria 9. Hungary 10. Italy . 11. Chile 12. Ceylon 13. United States* {Coloured) '''1'4. Portugal India, 1881 „ 1891 755 748 7^5 717 688 676 554 521 408 377 281 269 254 250 91 109 755 725 706 711 667 691 501 467 283 286 221 29 217 108 4 6 * The census return for the United States includes onlj' those over 10 years old. The figures here given are therefore approxi- mate. female population of the States, advance a step, whilst New South Wales and our .American cousins go down to that extent. India does not approach within one half of the literacy of even Portugal, so far as its males are considered, and the jproportion of its literate females is only just above a fifth of that in Ceylon, and this last, again, does rot reach a quarter of that of Portugal, or a seventh of that amongst the negroes and mulattoes of the Western Hemisphere. In Table G., that follows, are shown the component parts of the general total given above, together with the variations in the proportions that have taken place during the last decade : — Table G. — Provincial Statement of Illiterate and others. Province or State, &c. Ajmer - Assam Bengal Berar Bombay Sindh Lower Burma - .Upper Burma pentral Provinces Coorg Madras N.-W. Provinces t ouaii ■ - Pan jab ' Total, British Provinces Haidrabad Baroda ^Mysore Bombay States - Benp.al States N.-W. Province States Panjab Sfcites Total, States India - I 78888, Males. No. per 1,000 Males of- Illiterate. 1881. 1891. Literate, &c. 1881. 1891. Variation in Literates in 1891. 879 867 121 133 953 924 47 76 913 892 87 108 938 916 62 84 883 860 117 140 921 915 79 85 539 557 461 443 — .538 — 462 953 941 47 59 869 844 131 156 862 851 138 149 940 937 60 . 63 948 942 52 58 937 926 63 74 — 889 — iir 937 928 63 ■ ■ 72 894 856 106 144 888 895 112 105 900 887 100 113 965 945 35 55 963 965 37 35 947 941 53 59 — 902 — 98 909 891 91 109 + 12 + 29 + 21 + 22 + 23 + 6 -18 + 12 + 25 + 11 .+ 3 + 6 + 11 + 9 + 38 - 7 + 13 + 20 — 2 + 6 + 18 Females. No. per 1,000 Females of— Illiterate. 1881. 1891. Literate, &c. 1881. 1891 Variation in Literates in 1891. 994 999 997 999 993 995 904 998 990 991 998 999 998 999 998 996 998 999 996 99!) 996 992 997 996 998 99b 995 962 985 998 986 990 997 998 997 994 997 995 993 995 998 999 999 993 994 6 8 1 3 3 4 1 2 T 10 5 5 36 38 — 15 2 2 10 14 9 10 2 3 1 2 2 3 — 6 1 3 2 5 4 2 1 4 1 1 — t 4 6 + 2 + 2 + 1 + 1 + 3 + 4 + 1 + 1 + 1 + 1 + 2 H-3 + 3 + 3 + 1 -3 + 2 De 218 These last are on the side of improvement in every case save those of Lower Burma, Mysore, and the States under the North- West Province Grovernment. As regards the first, the Superintendent finds that in 1881 many families returned as under instruction the boys whom it was intended to send to the monasteries but who had not yet joined, and this misconception accounts for the relative falling off in the number of learners of that sex. The Mysore decrease also occurs amongst the pupils, and is confined to the western division of the State, where it affects the return of both sexes. No explanation is given of the fact in the report, and in the case of the small States of the north no report has yet been submitted. It may be noticed that the ratio of the not-illiterate amongst the females is higher in the States in 1891 than in the Provinces, and this seems to be due mainly to the extraordinary number of this sex returned as literate or under instruction in the Malabar States, which contain nearly one-half of the whole tale of the pupils of feudatory India, and more than that proportion of the literates. As these States were not called upon in 1881 to return literacy, they have been excluded from the table under discussion, but will be found duly entered in the next. The literate males are in considerably higher proportion in the Provinces than in the States at the present census. Upper Burma heads the list with nearly half of this Sex not illiterate. In its companion division, the monastery is less frequent, but it will be seen that the prdvision for female instruction, which is the accompaniment of British rule there, shows a much higher ratio than in any other part of the country. The Superintendent notes, with reference to the Phongi, or Buddhist monk, that there is a tendency to mddify the course of instructioiji in their Kyaung, so as to include the European method of teaching such subjects-as mathematics, a fact which indicates anything but the decline of their influence that was feared by some of the State educational authorities. In India proper it will be noted that more than 10 per cent, of the male community are provided with some stock of instruction in Madras, Bombay, Baroda, A.jm^r, Coorg, Mysore, and Bengal. Madras and Baroda head the list, the former having inherited tliis position, the latter having acquired it during the last 10 years. Upper India falls considerably below the average of either the Provinces or the country as a whole. Even Assam exceeds the North- West and the Panjab, which have only the Central Provinces between them and the bottom of the list. Assam, too, is distinguished by the highest relative increase of its literates, except that in Baroda. The latter lies chiefly in the Grujardth division of Western India, which has always been the foremost in the Presidency in the matter of education, the Broach and Surat districts coming not far short of Bombay City in the proportion of their literates. The ratio in the city of Calcutta is almost the same as that in Bombay, whilst in Madras, where the unskilled labourer finds less attraction, and educational institutions flourish, holds by far the highest place in this respect. In the Bengal Province it seems that the tract surrounding Calcutta is throughout the best supplied with literates, and that in the 'Eastern Division it is the dead weight of the Musalman lower classes that pulls down the average. Bihar, were it not for the hill tracts of Chutia Nagpur, would hold the lowest place in the whole province, whilst Orissa and the north occupy about the middle. In the Madras Presidency, Tanjore, the great Brahman centre, shows the highest ratio of literate males, but takes a comparatively low place as regards the instruction of its female population. The extreme south and Malabar are the only two districts besides Tanjore that contain less than 80 per cent, of illiterate males. In the Panjab there is great diversity in the proportion of the literate, and, on the whole, the lowest is found in the Trans-Indus districts of Bannu, Kohat, and Hazara, and on the borders of Eajputana. The western tract of the Central Provinces, bordering on Berar and Bombay, have the largest proportion of literate, and, excepting Jabalpur, the rest of the province gets more illiterate in proportion to the amount of hilly country included in it, until in the Mandla district there are but 29 males per 1,000 literate or under instruction. The details for the North- West Provinces and Oudh are not available for review in the present work. Berar, like Bengal and Bombay, has made considerable progress in the extension of literacy during the 10 years, and nearly reaches Sindh, where, as in Oudh, the progress has apparently been very slow. In the North- West Provinces there seems to have been practically none, as the per-centage of increase is but 3. It is possible, however, that there may be some special explanation of this peculiarity, which will be found in the local report, when published. The next table deals with the population divided into certain age-periods, as in Table B. On this occasion, however, it has been thought worth while to subtra'ct the 219 infant population, so as to leave a balance of possible school-goers. The result is a Literacy by difference of 2 per cent, in favour of the literates. In Upper Burma, for instance, the ^S^. illiterate' are reduced by 7,1 per 1,000 ; in Madras, by 26 ; in the Malabar States by 34, Table H.— The not-illiteratfe' pfer 1,000 of ieach Sex, by Age-period. Total. Total over 5. Province or State, &e. Ajmer Assam Bengal - Berav J Bombay - - 1^ Sindh J Upper Burma \ Lower Burma Central Provinces Coorg Madras fN.-W. Provinces \Oudh Panjab Total, Provinces Haidrabad Baroda Mysore - Bombay States Madras States Central Province States - Bengal States Jf.-W. Province State's Panjab States Total, States India 133 76" 108 84 140 85 462 443 59 156 149 63 58 74 72 144 105 113 239 22 55 35 59 109 3 4 2 10 5 15 38 2 14 10 3 2 3 3 5 7 5 35 1 2 1 1 154 89 125 97 163 100 533 510 66 175 175 73 67 i3o 83 161 113 133 273 21 64 40 ..66 127 9 4 5 2 12 6 18 45 ■ 2 20 13 3 5-14. 15-24. 25 and over. 7 . 9 C 40 1 .,2 1 1 97 66 96 104 149 76 .315 290 54 195 143 41 34 60 95 61 128 no no 209 17 35 22 30 79 93 . pq o s 3 ^ 9 6 5 3 17 9 15 44 4 37 18 2 1 4 4 13 14 8 49 1 2 1 1 191 12 172 114 5 95 156 7 13£ . 135 3 83 194 14 159 119 10 108 592 26 62C 584 67 602 99 3 64 190 20 162 196 15 185 86 4 84 79 2 81 97 4 125 i55 9 141 97 3 83 181 7 161 130 8 113 154 8 133 298 43 273 36 4 21 76 4 64 55 1 40 73 2 ■ 66, 129 9 124 151 9 138 3 4 1 4 5 15 34 2 10 8 2 2 3 3 7 9 6 40 1 2 1 1 and in th& Central Provinces and the Panjab by 7 only. The difference, in fact, varies with the, preyalpnce of,, instruction aimiongst the younger generation. Another feature in the return brought out by the distribution by age is that, except in Burma, the pi^portion of the literate males diminishes after the twenty-fifth year, -\vhilst it every- where increases from the first period to the second.^ The inference to be drawn is obviously that instruction in India proper is but a plant of tender years, whilst across the Bay of Bengal it dates from some generations. The diffusion of instruction through the community is appreciated more accurately Literacy by when we haye before us the relative extent to which it pervades each of the groups of caste or race, castes or races under which the population Av^as, classified in the preceding chapter. In the proAdncial volumes this return will be found under the title C, at the end of the Iniperial series. In the general series, to which the present review relates, it is numbered XVIII., and will be found in the second volume of tables. In the following proportional abstract, Table J., the leading features are given. It will be noticed that the ratio of the group ta the total population is not always identical with that given in the last chapter, partly for the reason stated in the preliminary note to Table XVII., namely, that two Provinces sent in their returns so far in arrears of the rest that the latter had been worked into their proportionate form in anticipation of the iijcprporation of the former. The reclassification therefore that seemed, advisable iii " the case of some of the castes found chiefly in Bengal and the North- West Provinces, was made only in the portion of the caste returns that waS corhparalively simple, involving, that is, merely re-addition, irrjespective- of further computation. Then, again, the return un^er discussion, omits the 25,0P0,00Q„or so not returning statistics of literacy. I ; Ee 2 220 Table J. — Literacy by Caste. A. Both Sexes. B. Mai,eb. C. LiTEBATJJS KNOWING Eng- lish. Per-centage of Per-centage of Per-centage of the Caste Group. Each Group on Total Popula- tion. Literate in each Group. Each Group on Total Males. Literate in each Group. English-knowing in each Group on On Total Literate. On Total Popula- tion of the Group. On Total Literate Males. On Total Males of the Group. Total English- knowing Literates. V- Total Literates of the Group. 1. 2. 3. 4. 5. 6. 7. 8. 9. I. Military Agriculturists a-46 8-10 3-94 9-78 8-01 7-07 2-82 1-55 II. Other Agriculturists 17-48 10-89 2-8l 17-43 11-12 5-5i 5-83 2-39, III. Cattle Graziers, &c. 4-20 1-01 i-ig 4-20 1-12 2-29 0-65 2-65 IV. Field Labourers 3-U 1-25 1-86 3-02 1-28 3-66 0-44 I-S5 V. Forest Tribes 5-04 0-44 0-40 4-95 0-45 0-78 0-20 2-01 Class A. Agricultural 39 29 21-69 255 39-38 21-98 4-81 9-94 2-03 VI. Priests 4'90. 16-81 15-78 4-96 16-98 29-52 20-29 5-37 VII. Devotees, &c. 0-85 1-00 S-41 0-87 1-01 10-02 0-31 1-38 VIII. Temple Servants 0-11 0-25 IO-32 0-11 0-23 18-11 0-07 i-3o IX. Genealogists 0-16 0-20 5-71 0-16 0-20 I0-9I 0-04 I -00 X. Writers 0'94 4-74 23-37 0-92 4-66 43-59 9-13 8-58 XI. Astrologers and Herbalists, &c. 0-10 0-39 17-58 0-10 0-34 29-72 0-97 11-28 XII. Ballad Reciters and Musicians 0'20 0-06 1-36 0-20 0-06 2-Sl 0-01 0-53 XIII. Singers and Dancers 0-05 0-06 6-62 0-04 0-03 7-62 0-02 1-19 XIV. Mimes, &c. 0-01 — 3-87 0-01 0-01 7-48 — 0-19 Class B. Professional 7-32 23-61 14-80 7-37 23-52 27-53 30-84 5-84 XV. Traders 4-11 13-74 15-38 4-15 14-09 29-30 6-6e 2-14 XVL Pedlars 0-05 0-01 1-42 0-04 0-01 2- 80 — 0-65 XVII. Carriers by Pack Animals 0-84 0-06 0-81 0-36 0-06 i-5i 0-01 0-84 Class C. Commercial 4-50 13-81 1413 4-55 1416 26-86 6-61 2-13 XVIII. Goldsmiths, &c. 0-57 1-20 9-69 0-57 1-24 18-69 0-26 0-97 XIX. Barbers 1-28 0-70 2-5i 1-28 0-71 4-80 0-33 2-14 XX. Blacksmiths 0-92 0-50 2-53 0-92 0-52 4-82 0-21 1-87 XXI. Carpenters and Masons 11.3 0-97 3-97 1-14 1-00 7-61 0-24 I -10 XXII. Brass and Coppersmiths O'U 0-20 8-43 0-11 0-21 16-34 0-08 1-86 XXIII. Tailor.s 0-23 0-23 4-64 0-23 0-24 8-93 0-14 2-73 XXIV. Grain Parchers, &c. 0'53 0-38 3-29 0-53 0-39 6-46 017 1-97 XXV. Betel Leaf Sellers, &c. 0-09 0-13 6-go 0-09 0-14 13-23 0-12 4-12 XXVI. Weavers and Dyers 3-lC 2-09 3-o3 3-15 2-13 5-83 0-90 .-93 XXVII. Washermen 1'02 0-24 I -08 1-01 0-24 2-08 0-12 2-3o XXVIII. Cotton Cleaners 0-30 0-04 0-64 0-30 0-04 1-25 0-01 0-77 XXIX. Shepherds and Blanket Weavers 1-78 0-39 I -02 1-77 0-41 1-98 0-20 2-28 XXX. Oil Pressers 1-66 1-16 3-22 1-64 1-20 6-32 0-57 2-17 XXXI. Potters 1-14 0-34 1-37 1-15 0-35 2-62 0-26 3-38 XXXII. Glass and Lac Workers 0-05 0-03 2-56 0-05 0-03 S-o5 0-04 5-10 221 Table J. — Literacy by Caste — continued. A. Both Sexes. I i. Males. C. LiTEBAIES KNOWING Eng- lish. Per -centage of Per Each Group -centage of Per-centage of the Caste Group. Each Group on Literate in each Group. Literate in each Group. English-knowing in each Group on On 1 Total Popula- tion. On Total Literate. Total Popula- tion of the Group. on Total Males. On Total Literate Males. T°tol Total Total Males ^°S''?'^- I^itf^f ^s of the ^]l°^fe f the Group. I-iterates. Group. 1. 2. 3. 4. 5. 6. 7. 8. 1 9. XXXIII. Salt and Lime Workers 0-54 0'15 I '29 0-53 0-15 2-49 0-03 0-76 XXXIV. Goldsmiths' Refuse Cleaners - — — 1-89 — — 3-90 — — XXXV. Iron Smelters and Gold Washers 001 — 0-70 0-01 — 1-35 — 0-57 XXXVI. Fishermen, &e. 3-17 0-80 i-i3 3-12 0-79 2-17 0-36 2-o5 XXXVII. Servants, &c. 0-07 0-02 1-43 0-07 0-02 2-75 0-01 I -14 XXXVIII. Distillers and Toddy Drawers 1-84 2-44 6-II 1-80 2-47 11-87 0-74 I.- 33 XXXIX. Butchers 0-20 0-03 0-63 0-20 0-03 I -20 0-01 • 1-64 XL. Leather Workers 4-59 0-53 0-S3 4-58 0-55 1-04 0-13 i-i3 XLI. Watchmen and Village menials 4-68 0-63 0-62 4-59 0-64 I-2I 0-44 3-09 XLII. Scavengers 1-32 0-17 0-59 1-35 0-17 I'll 0-12 3-25 Class D. Artisans, &c. 30-39 13-47 2-02 30-19 13-70 3-91 5-49 1-83 XLIII. Grindstone Makers 0-01 — o-5i 0-01 — 0-98 — — XLIV. Earth Workers and Quarrymen 0-42 0-03 0-33 0-41 0-03 0-62 0-01 I -So XLV. Knife Grinders — — 1-22 — — 2-29 — 1-97 XL VI. Mat and Cane Workers 0-22 0-03 0-61 0-22 0-03 i-i6 0-01 0-79 XLVII. Hunters and Fowlers 0-28 2-04 0-79 0-28 0-05 1-52 0-02 1-56 XLVIII. Miscellaneous Vagrants 0-12 0-03 I-I9 0-13 0-03 2-i5 0-01 i-3i XLIX. Jugglers and Acrohats 0-10 003 1-29 0-09 0-02 2-24 0-03 4-32 Class E. Vagrants 115 016 0-67 1-14 0-16 1-26 0-08 1-87 L. Musalmans of Foreign Titles 12-52 9-46 3-48 12-54 9-56 6-58 6-19 2-91 LI. Thibetan and Nipali Tribes 0-03 0-10 5-64 0-09 0-10 10-25 0-07 3-09 LII. Burmese Tribes and Chinese 2 '79 12-53 20 -68 2-72 12-46 39-53 0-58 0-20 LIII. Western Asiatic Races 0-04 0-42 45-50 0-04 0-26 54-89 2-98 33-o8 LIV. Mixed Burmese Tribes - 0-01 0-01 7-99 0-01 0-01 15-38 — 1-64 LV. Indefinite Indian Castes 1-08 1-40 5-94 l-ll 1-37 10-69 1-24 3-95 LVl. Europeans 0-06 0-98 73-24 0-09 0-81 78-67 21-06 95-79 LVII. Eurasians ■ 0'03 0-35 52- 18 0-03 0-19 54-73 7-16 91-27 LVIII. Native Christians 0-72 2-05 13-07 0-72 1-67 20 -o5 7-38 i6-2S JjiX. Goanese, &c. 0-01 0-05 i8-6i 0-01 0-04 26-61 0-37 26-18 LX. Africans O'Ol 0-01 3-i8 0-01 O-Ol 5-3o 0-01 4-98 Class F. Races and Miscel- 17-35 27-36 7-25 17-37 26-48 13-17 47-04 7-67 laneous. Be 3 !222 Concentra- tion of literacy in certain groups. Table J. deals with the literate, including those still under instruction, in three sets of ratios. First come the proportions respectively borne to the total population by each of the caste groups. In connection with this is the series showing the proportion of the literate in each group to the total literature, together with a second series, in which is given the ratio between the literate in the group to the total population coming under that group. The second set of proportions deals with the males only, and is added merely on account 'of the; practically universal illiteracy of the other sexi. It comprises the same subjects as the preceding set. The third and last set indicates, first, the distribution by groups of' the literate persons pf both sexes who know English, and, sefcondly, the ratio bqtwqen the latter and the total literate of the group. » This detail is so far interesting, since instruction above the primary limits is practically conducted in English in all the public schools and colleges. It will be treated, accordingly, along with literacy in .general. The following is the way in which the table under discussion is intended to be used : — Group II., Other Agriculturists, contains 17'48 per cent, of the population, with 10-89 per cent, of the literates, the latter bearing a ratio of 2'8.1 per cent, on the population of the group. Amongst the literates are 5 "83 per cent, of the total body of English-knowers, and these, in turn, form 2"39 per cent, of the literates in the group, so that in every 10,000 six know that language, or one in 1,667. With this explanation the figures may be left to speak for themselves, so far as the details are concerned, and it is worth while to bring to notice here only the more prominent features in this curious return. For instance, if both sexes be taken together, as in the first section of the table, it will be seen that in 11 groups only are the literate as high as 10 per cent, on the included population. The marginal ox- Table K. tract reproduces the information regarding these 11. They com- prise just under 14 per cent, of the population, just over half the literate population, and more than three-fourths of those who can read and write English. If the collection be re-grouped into more minute sections, it will be seen that the Brdhmans, "Writers, Traders, Native Christians, Temple Servants and Herbalists, &c., who constitute the strictly native portion of the whole, contain 11 per cent, of the popu- lation, 38 of the literate, and 45 of the English-knowers. The Burmese and Parsis with the few Armenians and Jews come next, with 2-8 per cent, of the popu- lation, nearly 13 of the literate, and just above 3^ per cent, of those who know English. Finally, we have the European and Eurasian element, which accounts for just under one in a thousand of the population, 13 in the same number of the literate, and 282 of the English-knowing part of the community. Oatside this circle is found about 23 per cent, of the latter population, or about the same proportion as is contributed by the Europeans and Parsis, &c., taken together. It will also be noticed that the Br^hmans, Writers, and Europeans monopolise more than half of this class of the literate, and the Traders' Eurasians, and Native Christians, a fifth more. As regards the introduction of the Herbalist and Astrologer, it should be explained that the former is apparently one of the best instructed classes in Eastern Bengal, to which part of the country he is, as a separate caste, confined. The Temple Servant group, again, owes its position to the Satani of Madras and Mysore, where this class is most prevalent. We may now turn from the general section of the table to that which treats of males only. Here we find that no less than 20 of the 60 groups returns 10 per cent. Caste Group or Eace. Per-centage on total of Population. Literates. English- knowing Literates. 1. Priests 4-90 16 81 ■20-29 2. Temple Servants 0-11 25 0-07 3. Writers 0-94 4 74 9-13 4, Herbalists, &c. 0-10 39 0-97 5. Traders 4-11 13 74 6-60 6. Burmese 2-79 12 53 0-68 7. Parsis, &c. 0-04 42 2-98 8. Europeans - - . 0-06 98 21-06 9. Eurasians 0-03 35 7-16 10. N.ativp Christians 0-72 2 05 7-38 11. Goanese Christians 0-01 05 0-37 Total - 13-81 52-31 76-59 223 and over of literates in its community. Table L. Caste Group, &c. Per-centage on totals of Males. Literate Males. English- knowing Literate Males. The additions to the former list are the ' Devotees, Genealogists,* Groldsmiths, Brass-smiths, Betel Leaf Sellers, Distillers, Nipali and Thibetan tribes, and the mixed races of Burma, with the group that had to be set apart for indefinite entries, containing a good number of the writing castes serving at a distance from their native province, and thas entered under some misconstrued title. These additions enlarge the scope of the collection considerably. Instead of 14 per cent, of the population we get over 18 of the males, with 58^ per cent, of the literate of that sex and 79^ per cent, of those who know English. The groups in which female instruction is more prevalent take, of course, a lower place in this statement than the last. This remark applies to the Writers, Temple Servants, Herbalists &c., Parsis &c., Burmese, Europeans, Eurasians, Native Christians and Goanese, to all of the former selec« tion, in fact, except to Traders and Brahmans. In the case of the former, there is no doubt that some of the difference is attributable to the number of literate men who come from Rajputana, &c., to the centres of commerce in British territory without their families, but more to the general cause, namely, apathy, as in the case of the Brahman. If, again, we take as the standard proportion the ratio of the literate for all India, which we have seen in the beginning of this chapter to be 8-66 for males alone, our roll is increased by one name only, that of the Tailors ; but by including the fair sex, the Tailor falls out, and the Nach girl takes his place. It is interesting to compare a few castes selected as the least illiterate of each Province, as is done in the following table : — Table M. 1. Priests 2. Devotees 3. Temple Servants Genealogists Writers Herbalists, &c. - Traders Goldsmiths Brass-smiths Betel Leaf Sellers Distillers, &c. - Thibetans Burmese Burmese, Mixed Parsis, &c. Indefinite Hidians 17. Europeans 18. Eurasians 19. Native Christians 20. Groanese Christians Total 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 4-96 0-87 O'll 0-16 0-92 0-10 4-15 0-57 0-11 0-09 1-80 0-09 2-72 0-01 0-04 1-11 0-09 0-03 0-72 0-01 18-66 16 1 4 6 14 1 2 12 1 1 58 •98 20 •01 0^ •23 0- •20 0- •66 9^ ■34 0^ •09 6- •24 0- •21 0^ •14 0^ 47 0- •10 0^ •46 0^ •01 •26 2- •37 I- ■81 21 ■ •19 7- •67 1- 04 0- 48 79- •29 •31 •07 •04 •13 •97 •60 •26 •08 •12 •74 •07 •58 ■98 ■24 ■06 •16 ■38 ■37 ■45 Literate per 1,000 Literate per 1,000 of Sex. of Sex. Caste, &e. Group. Caste, &c. Group. Males. Females. Males. Females. Madras : Mysore : Brahman Kannakkan Karnam VI. X. X. 722 658 587 37 21 13 Brahman Bania Komti VI. XV. XV. 678 819 664 38 12 Komti XV. 605 9 Central Provinces : Nayar - I. 490 125 Brahman VI. 317 7 Native Christians - LVIII. 218 76 Kayasth X. 475 i 16 Eurasians LVII. 786 720 Vidhiir - X. 342 4 Bania XV. 388 — Bombay : Haidrabdd : Brahman K£yasth VI. X. 645 687 33 212 Brahman Kayasth Vidhur VI. X. XV. 569 438 151 5 Prabhu - X. 791 164 Brahma Kshatria - X. 648 268 Baroda : Bania Shrimali - XV. 697 15 Brahman VI. 559 24 Bania, Unspecified XV. 518 18 Prabhu Bania X. X. 790 776 87 Berar : Panjib : BhAbra XV. 463 7 Brahman Kayasth 'Prahhu - ViVI. X. X. 638 557 764 21 160 160 Ardra - Bania Khatri XV. XV. XV. 381 429 394 6 3 7 .Vidhdr X. 368 — Sud - - - XV. 416 8 Bania XV, 480 — Brahman VI. 191 2 Kdmti XV. 349 — K4141 XXXVIIL 164 5 Bengal : Kayasth X. 434 68 Saiad ■ L. 110 6 Brahman VI. 477 23 N.-W. Provinces : Kayasth X. 556 41 Brdhman* VI. 180 5 i Baidya - XI. . 734 139 Kayasth X. 610 29 Karan - - - - X. 604 16 Bania - XV. 265 4 Rinia - XV. 280 4 Bania Agarwfil - XV. 468 6 * The Brahmans in the N.-W. Provinces are largely agriculturists. Be 4 224 English- knowing literates. Quality of IhevsLcy. The final computation made above brings us to the fact that in India, as a whole the very moderate average of 46 literate persons in a thousand, is not attained by 81-35 per cent, of the population, but is the result of greater prevalence of instruction amongst the remaining 18-65. In the case of the males alone, the standard rises to 87 per 1,000, but it is not reached by more than 18-89 per cent, of the sex leaving 81-11 below it. ^ The return of those who know English shows a ratio of 4-4 per cent, on the tota literate. We must subtract, however, the Europeans and Eurasians from the account which then amounts to 3-2 only, or 1-4 in every thousand of the community. From' the detailed table 'it will be seen that, excluding the Europeans, Eurasians, Nipali Africans, and Parsis, the latter proportion to the literates of the group is' achieved only in the case of the Brahnaans, Writers and Herbalists, with the group of the indefinite castes. There are, it is true, four or five other groups that show a per-centage in slight excess of this, but they are all chiefly recruited from Bengal where this part of the enumeration seems to have been unsatisfactory, since nowhere else do we find the Scavenger, Potter and Acrobat in such exalted company. The entire number returned as knowing English, including Europeans and Eurasians, was 537,811, or 386,032, if the foreign element be excluded. This, too, includes a certain proportion of those who are not yet emancipated from their studies, as has been already- remarked in the beginning of the chapter. Some of the Superintendents, on the other hand, seem to think that the return includes, from excess of caution, only those who habitually use English in their daily life, and not the numerous class that learn a certain amount of that language at school, but carry the use of it no further than the last examination before their escape from that stage, and cease to be able to read and write it after the lapse of a few years. The census return seems to compare but poorly with the Departmental record in this respect, for the latter gives an average number of pupils studying English of 290,741 per annum during the last decade, beginning with 187,420, and ending with 353,515. The average period of study is not accurately known, but one would have expected to find at least 700,000 or 800 000 of the above number amongst the English-knowing literates. But apparently' the study of English ends in a very rudimentary stage, for with an average annual attendance of nearly 337,000, studying in that language for the last five years, only 15,200 presented themselves for the matriculation examination at the Universities, or 70,000 during- the whole period. As English is the language of instruction at the colleges aflfiliated to the latter institutions, it is presumably an important subject at the matriculation test if not the most important. But we find from the same returns from which the above quotations are made, that the ratio of the successful for the five years in question was 47-74 in Calcutta, 26-87 in Madras, and 25-41 in Bombay. The other (Jniversities need not be counted, as they are, comparatively speaking, in their infancy. But at any rate the out-turn of 25,680 in five years of youths up to matriculation' standard even with the possible successes under the sixth standard elsewhere, are scarcelv results that need make the census returns blush in comparison. It is interesting, though not arising out of the census returns, to carry this subject a little further, and now that we have seen the concentration of literacy in considerably under a fifth of the population, including foreigners, to give a passing glance at the quality of that acquirement. For this we must resort to the Departmental tables, of which a proportional abstract is given in the margin. To start at the bottom, are 922 pupils in every thousand, in the primary stage, of whom 825 are classed as "elementary" and the rest as " upper." The proportion 'in the lower section that has not yet o-ot into printed books is not complete, owing to the introduction of the private schools, where the information is not tabulated, but in those under the State it seems that in the total lower primary section 30 per cent, of the boys and 361 of the girls are in that stage. The ratio varies from 3-6 in the Panjab, where the Kuran schools and the like are not in the returns, to 87 per cent, in Upper Burma. Amongst tje girls, omitting the Panjdb, Assam comes lowest, with 15, and Berar highest, with 85 per cent. Jt is remarkable that Table K. Per 1,000 Pupils. University 4 N"ormal Institutions 1 Technical Institutions - 4 High Schools 15 Middle Schools - .54 Upper Primary Schools 97 -r f Reading Books ^Lower 1 j^^^ Reading Books - rrimary ^Not stated 492 220 113 Total Pupils 1,000 225 in the last province and its neighbour to the east, the pupils of each sex not yet up to reading books outnumber those in the advanced stages of instAiction. The same table shows the proportion of the Ipwer primary, scholars to the total by provinces. It is unnecessary to reproduce it here, but it may be mentioned that in the case of boys, the average is 82 per cent., and Assam heads the list, if we omit Upper Burma, with 90-4), and Bombay comes lowest, with 65^. The figures for girls, averaging 91|- per cent., rise to 99 in Assam, and 'fall to 83 in Berar. Out of the thousand, we have left, after the primary scholars, 78 pupils in higher stages. There are 69 in the Middle and High Schools, and four in Technical Institutions, most of which are not specified in the return by the object of their course of training. Normal schoools and colleges account for one in the thousand, leaving four for the University. It is worth while to follow the progress of the candidates for matriculation a Universities, little further than was done above. ' It was there shown that an average of 15,200 per annum had presented themselves during the last five years, and that just under 34 per cent, had passed. The acconapanying Table N. shows that out of every hundred that were examined in Table 0. — Results of University Examinations for the Calcutta 52 failed ; in five years 1886-87—1890-91. Madras, 73, and in Bombay, 74. In the next section of the table we pursue the successful only, and find that out of every hundred of them, 36 in Calcutta, 2 in Bombay, and 48 in Madras did not attempt to graduate. Of the remaining 64 that came up for a degree in Calcutta, 29 got one. In Bombaj'-, out of 98, 46 were successful, and in Madras, the 52 that tried came out with 26. In the case of Bombay, there is evidently some change of date or system to account for the relative scarcity of those who were content with matriculating, and there may be, in like manner, some special cause for the astonishing propor- tion of this class in Madras. But so far as the figures go, it seems that out of the four pupils per mille that go up for matriculation, those that get a degree are 1 in 10. The lower test, known as the First Arts, is more popular, and even failure at it is often considered as a certain qualification for salaried employment but the main points brought out by the figures published in these returns seem to be, first, the insignificant number of pupils that carry instruction beyond the rudiments ; secondly, the remarkably unprepared state in which the minute remainder appear for ^latriculation ; and lastly, the relatively infinitesimal number that obtain a University degree either in Arts, which is the faculty most favoured, or_in more special subjects, i- 78388. Y f Universities. Calcutta. Madras. Bombay. A.: — JtJxamined for Matriculation* 100 00 100-00 100 00 r Failed \ Passed . 52 '26 : 73-13 74-59 47 •74. 26-87 25-41 Examined for Degree in Arts .20 -59 10-37 12-87 „ » Law . 7-06 1-80 2-44 „ „ Medicine 2i-30 ■ 1-52 6-66 „ „ Civil Engineering '0-55 0-16 2-97 Total examined for Degrees 3o-5o 13-85 24 "94 j Passed 13-97 6-80 11-78 [ Failed 16-53 7-05 J3-16 Total not appearing for Degrees 17-24 l3-02 0-47 B. — Passed the Matriculation 100 00 100 00 100 00 Examined for Degree in Arts 43-13 38-61 50-66 „ n Lav/ 14-79 6-68 9-62 J, „ Medicine 4-83 5-64 26-21 „ Civil Engineering 1-15 0-61 11-67 Total examined for Degree 63-90 5i -54 98- 16 j Passed . - - - I Failed - 29-25 25-32 46-35 34 -05 26-22 51-81 Total not appearing for Degrees - 36- 10 48-46 >• 1-84 C, — Per-centage of those examined who passed 38-60 55-32 47-09 for Degree in Arts. Per-centage of those examined who passed 63-93 28-96 33-79 for Degree in Law. Per-centage of those examined who passed 53-78 32-44 48-27 for Degree in Medicine. Per-centage of those examined who passed 47-86 32-14 53-88 for Degree in Civil Engineering. Per-centage of those examined for 45-77 49 '34 47-22 any Degree who passed. » The actual figures are, Calcutta, 21,238 ; Madras, 34,393, and Bombay, 14,774. Books Province. published in 1890-91. Madras 1,022 ; Bombay 2,044 ' Bengal 1,225 ! W.-W.-.Provinces - 1,107 Burma ... 149 Assam - - - - 22 Berar . - - 13 Central Provinces 13 Total - 5,595 a number which amounted, in a year of census, to one person in about 213,000 of the population to whom the career was open. Literature. F.n the chapter on occupation it was shown how stiiall a fraction lived by literature, and though the annual returns show an imposing array of publications, the review of the literary activity of the year by the Official Reporter is rather discouraging reading. According to this authority, a few works on Sanskrit texts, with an occasional drama on a historical occurrence or a subject of the day are all that are likely to survive th'e year of their birth. A good deal of this infant mortality, so to speak, seems to be attributable to the very high proportion of the publications which deal with the text books prescribed' for University or School examinations, or other ephemeral works designed for the same market. The most striking characteristic of the out- turn seems to be the absence of originality in list does not want variety, as will be seen from the marginal statement of subjects, with, of course, the qualification that rather over a third are translations or republications. The language in which the works are issued also is a matter not devoid of interest, and it appears that in English 660 were published, with 955 in polyglot, 2,157 in a vernacular tongue, and 424 in the three Oriental classical languages. But a more favourate outlet for budding talent is found in journalism, of which we find 490 exponents in the list. The largest circulation is stated to be 20,000 in the case of one paper in Bengal; about 6,000 is the maximum in Bonabay, and 5,000 in Madras. Elsewhere it seems to rarely reach a thousand. This does not represent, of course, nearly the number of readers, for the economical practice of private circulation, or of perusual at cheap libraries, is far more extended in India than in many other countries. But lithography and disregard for typographical appearance enables an enterprising publicist to start a local broadsheet at a very small' cost, and what with the restrictions of career imposed upon themselves by a solely literary caste or two, no country, probably, has more represent^itives than India of the hero of the Romaic ballad : — &€plJi.hs l in 188-1 th€| reborn seems to have been VQry de^cient, the increase was suc]]i as to induce t|ie Superintendent to Jtiave a special iavestigation made, the results of w-hich, it is ^-ratifying to -see, confij:',i?je5" th-e acdtH'afcy of the Ff2 '-■ " 228 present figures. In Bombay, too, tlie dift'erence between the two years was enough to justify inquiry, with the same result. But, even though we admit the accuracy of the statistics, it is to be feared that the enumeration is prone in too many cases to fail in the matter of diagnosis. For example, in a country where so much disease is put down to fever, there must be numbers of patients who are for a time, at least, " off their head," without being permanently deranged, and some of them aire no doubt returned at the census from ignorance on the part of their relatives of the temporary, character of the malady. In Europe, there is either medical authority for the case, or the latter is not returned hj the householder at all, and it is not Polonius alone, who is puzzled to give a definition of insanity. In the case of leprosy the difficulty of diagnosis is still greater, and, apart from the contusion between the true disease and leukoderma, it is not improbable that a good ieal of syphilitic sore is set down by the untrained observer to the former. It seems by no means certain that the confusion of leukoderma is not due to the wording of the instructions in former years, or to the misuse of the vernacular words for it by Eui;opean professipnals, for, whilst in the mouth of the rustic the two maladies are entirely distinct, to, the occidental mind there is always a connection between leprosy and whiteness, a casting back to Gehazi or other Biblical characters, and the patches of white seen on the dark skin of a native of India in cases of leukoderma are certainly very striking. It is probable that a good deal of the decrease in the number of lepers returned in 1891 from tracts in which they were formerly remarkably high, is nominal, and due to greater accuracy of diagnosis, and this being so, it is impossible to, say for certain whether this disease is, on the whole, stationary or not. In the provincial reports, . instances are given by the Superintendents of undoubted decrease,! but the assumption that this is tjhe case with the country at large is hardly justifiable at present. 'The qi^estion is one that will be referred to in a later portion of this chapter. The following statement shows in a proportional form the prevalence of the four infirmities in question, and, for the sake of comparison, the corresponding figures for some western countries are added. As i-egards the latter, the most striking points Table A. — Propoktional Summary, showing Prevalence of Infirmities per 10,000 of each Sex. Insane. Deaf Mutes. Blind. Lepers. Country. * Males. Females. Males. Females. Males. Females. Males. Females. India, 1881 4 3 10, 7 22 24 8-5 3 India, 1891 .3 2 9 6 16 17 6-8 2 England and Wales 31 33 6 5 9 8 — — Scotland 38 39 3 2 9 8 — — Ireland , , 38 . 34 8, 7 11 12 — — Italy — — 6 5 8 7 — — Austria 22 19 15 11 10 8 " Hungary 22 18 14 12 13 13 — — Portug«il 23 17 3 2 21 19 — — Ceylon 13 9 2 1 24 18 — — Victoria 37 31 4 3 11 6 — — New Zealand 33 22 2 2 4 3 — — United States , - ' ■ , ,' i 35 32 7 6 10 9 — 229 of. difference are, first, the relative freedom of India from insanity, and then the relative frequency m that country of blindness. The former affliction is very markedly prominent m the more advanced of the western countries -and in the British colonies in the south, (^eylon, for some reason or other not apparent, stands, in this respect, between east and west. The return for Italy seems to include a smaller class than the rest, so it has been omitted. Jt is also noticeable that the higher proportion of congenital deaf -mutism amongst males, that has been commented on in the English census, is maintained in India, but in the case of blindness, in Ireland alone, of the western countries mentioned, does the ratio amongst females exceed that in the other sex, as is found to be so often the case in the Bast. The proportion of deaf-mutes in Austria-Hungary appears from this return, which deals with the figures of 1880, to be higher than that which prevails in India, and in striking contrast to the low ratio in the rest of the European countries selected. Leprosy is returned in India alone, and here we find the proportions in the two sexes to differ more widely than in the case of any other of the infirmities under review, a point that will receive notice further down. The greater prevalence of insanity in western countries is easily accounted for by the difference of life there from what it is in India. We have seen that in the latter the brain-energy of the masses is not overtaxed by literary pursuits, and the struggle for life is not the competition between man and man so much as between man and the soil and sky. In the one case we have almost infinite variety, entailing continual adaptability and change of thought; in the other, monotonous devotion to a certain and lifelong task, based on previous experience. The man of the west runs more and more towards the busy novelty of the town, whilst his Oriental compeer placidly and contentedly chews the cud of village custom. Then, again, though we find that a certain proportion of the insanity of India is attributed by experts to the abuse of hemp drugs or opium smoke, alcoholisation has as yet made no way, except, perhaps, amongst some of the Central Hill tribes and a few of the literate class and the lower strata, who look upon this indulgence as a symbol of their emancipation from former prejudice. This fertile cause of insanity is thus absent. So far as to the men, and much the same arguments will apply to the other sex. It can be seen from the return under discussion how much nearer are the ratios of the two sexes in the "West than in the East. The busy household life of the middle class woman in a country w;here the employment of domestic servants is restricted to the highest families, the seclusion of the upper class within the four walls of the Zanana, away from all chance of active participation in the affairs of the outer world, and the continuous outdoor labour of the lower orders, contain little stimulus towards the over-taxation of the brain, nor does a school life that generally ends at 12 years old, if not earlier, foster any tendencies in the same direction. On the other hand, the prevalence of blindness in India to so much greater an extent than in more temperate latitudes, and amongst women more than amongst men, is a matter that seems to be due to more complicated causes. It has been said that, speaking generally, there is a tendency for this infirmity to increase in intensity as the equator is approached, though the case of Norway is cited in contradiction of the hypothesis, and prevalence of snow in wide expanse may have some influence on the sight in the extreme north ; but the returns for India certainly seem to bear out the notion that heat, accompanied by a dry atmosphere, is, other things being equal, inimical to sight. There are facts to be found in the tables that militate against this view, but they refer to restricted areas, where some special local influence may be operative. Taking the country as a whole, however, the hot and dry plain and plateau seem more favourable to the development of ophthalmic defect than the moister, though perhaps warmer, air of the coast and the purer atmosphere of the hill tracts. There seems to be nothing peculiar in connection with the prevalence of deaf -mutism in India. As in many other countries, there seems- to be a tendency for this defect to haunt the valleys of certain streams. For instance, the Ohenab, the Gandak, and the Makhua, all three Himalayan rivers, bear a specially evil repute for producing both deaf -mutism and cretinism, with the frequent accompaniment of goitre! In Burma the hill country shows a greater prevalence than the plains, and in Assam the same feature is prominent, and it is by no means improbable that the practice of consanguineous connection that is common amongst the tribes of those tracts has something to do with the spread of the infirmity, for beyond the Gangetic basin, in the Central Belt, the corresponding tribes show no difference from the population of the plain below. Leprosy is a disease with which the Bast is in the present day credited' to an extent beyond, perhaps, its deserts. Its prevalence will be examined in a later portion of this chapter. ' ■ Ff 3 230 We come now to the detailed figures regai'ding the four selected infirmities which will be found tabulated for each province and State, by sex and age, in Tables XII. to XV. of Volume I. of the Imperial series. The scrutiny of these data is a matter that must be left in great measure to professional experbs, and it is not for laymen to draw conclusions from them. It'*' is proposed, therefore, to exhibit the figures here in their proportional form, with a few words of comment or explanation on points immediately connected with the enumeration and kindred operations. The tables that follow are for the provinces only. In the first, marked B., is given the proportion amongst 10,000 of the afflicted found at each age-period, preceded by the corresponding figures for the population, as a whole, similarly distributed. Table B. — Showing the Distribution by Age of 10,000 under each Infirmity. Males. Females. Age-period. 1 1 1 si O., i at o 1 p '6 _B, S 281 c 0-4 1,416 166 457 427 44 1,524 170 528 81 5-9 1,433 .552 1,445 639 84 1,393 532 1,450 409 180 10-14 1,145 789 1,317 647 236 948 789 1,154 417 399 15-19 836 924 1,070 594 400 817 937 1,024 417 602 20-24 796 1,045 965 609 572 890 988 952 453 735 25-29 866 1,233 897 633 873 900 1,000 '859 518 924 30-34 827 1,272 823 662 1,205 833 1,108 801 621 1,186 35-39 613 975 602 563 1,212 556 885 544 541 990 40-44 626 995 620 693 1,516 , 616 980 622 738 1,292 45-49 369 573 374 565 996 328 608 367 598 788 50-54 404 561 452 735 1,174 421 727 484 868 1,010 55-59 183 287 242 626 495 . 174 328 252 711 465 60^nd over 469 600 729 2,601 1,185 583 911 955 3,422 1,335 Age not stated 17 28 7 ' 6 8 17 37 8 6 13 InsaDity. The mean age of the living, as gathered from the general age-tables, is, roughly speaking, 24 years, and if we divide tlie population shown under each infirmity at that age, we find the distribution to be as given in the margin. Nearly two-thirds of the insane in both sexes are over the age in question, the deaf mutes are nearly halved. In the case of the blind, there is more divergence between the two sexes, and the males begin to suffer from this affliction at an age considerably earlier than the other sex, whilst as regards leprosy, the reverse is the case, and 20 per cent, of the female lepers are under 25 years of age, as compared with no more than 13 per cent, of the males thus affected. In respect to the more detailed age-periods, it must be borne in mind that the tendency to select multiples of five in returning age, a feature by no means unknown in Europe, is very marked in an Indian census, so that, aw the more favoured multiples are the even ones, it is advisable to consider the tables in the decennial periods of 25-34, 35-44, and so on. Taken in this way, the figures show a gradual, rise in the case of males of unsound mind, from the fir.st five years ,of; life to the period ending at the 35th year, from which point insanity declines with remarkable rapidity. On / Males. I'emales. 0-24. 25 and over. 0-24. 25 and over. Insanity Deaf-mutism Blindness Li'piosy Total Population .348 525 292 134 563 652 475 708 866 437 342 511 198 200 557 658 489 802 800 443 281 Deaf- mutism. Blindness. the other side, young girls suffering from this defect are a trifle more numerous, relatively to the total number of the insane of their sex, but between 25 and 34 the proportion is less than that which rules at those ages amongst males, whilst after 45 it rises considerably above it. The deaf mutes exhibit a distribution more in accordance than the rest with that of the population as a whole, but the extremities of the return, especially in the case of the females, indicate, the inclusion of the deaf, at advanced ages. The first period, if the congenitally afflicted alone had been returned, .would have contained th^ largest number, and there would. bave been a gradual tailing off, as death, in the ordinary course of events, carried off its victims. It is not so, however, in tlie table, but tho maximum is found in the next period, that between 5 years and 10, and both sexes take a sudden rise after 55 years of age. There are obvious objections on the part of parents to admitting thab a , young child is deprived of both hearing and speech, especially before completion of the , age within which betrothal is possible, and this may account for a good deal of the inconsistency just mentioned. Blindness, in the case of females, is most decidedly the affliction of old age. The table shows that over 40 per cent, of those returning it are past 55, and the increase is continuous, though not regular, through life. In the case of males, there is more irregularity, owing, however, to some extent, to the predilection for returning 25 years, when the actual age is probably anything between 20 and 30. The proportion over 55 is only about a third, and the greater number found amongst the young, as compared with the other sex, is very remarkable. It has been remarked in several of the provincial reports that there was a frequent inclination to return as totally blind persons of advanced age whose sight was dimmed by glaukoma. On the other hand, a few cases were mentioned in which the enumerator, owing to misreading his instructions, tacked on the qualification of "from birth" to this infirmity, as well as to that to which it properly belonged. Leprosy, in point of distribution between the two main periods of life, corresponds. Leprosy, in the case of females, almost exactly with blindness, but in detail it is very different. There are more than thrice the number of blind girls of lender age than of lepers, the latter are still in a considerable minority between 5 and 15, but the marriageable age being past, they then begin to outnumber the others, relatively, of course, to the respective totals afflicted. The lepers, however, are not recruited, as are the blind, in old age, and their life seems a comparatively short one. Male lepers of less than five years old are still less prevalent than those of the other sex, and it is not till between 25 and 34 that they relatively outnumber the blind. Like the other sex, they maintain their comparative superiority until 55 years of age, when they fall to about the half. Even here, however, they seem to considerably outlast the insane and the deaf mutes. We may now review the age-return of the afflicted population from another standpoint, namely, in its relation, not to the total of the persons Suffering under the respective infirmities, but to that of the entire population of the same sex and age: This ratio is given in Table 0., below. Table 0. — Showing the number afflicted in 100,000 of each Age. Males. Females. Age- period. 1 P i S a* -1 a s 1 "a a a to 1^ 0-4 5-9 10-14 15-19 - 20-24 26-29 - 30-34 35-39, - 40-44 45-49 - 5M.4 55-59 - 60 an(J over - AH ages 86 186 240 310 337 341 385 444 509 600 666 939 1,285 367 4 13 23 37 44 48 > 52 54 54 53 47 33 4f 34 ■ 31 96 109 121 115 98 94 93 94 96 106 125 147 95 49 73 93 116 125 120 131 160 181 250 297 560 907 164 2 4 15 36 53 75 108 147 180 201 216 201 188 74 55 125 178 208 196 206 250 303 351 475 315 879 1,179 277 2 9 19 26 25 85 30 36 36 41 39 42 35 22 21 64 75 78 66 59 60 60 62 69 71 89 101 62 31 49 74 86 85 97 125 163 201 306 346 682 9$6 168 1 3 10 18 20 25 35 44 52 59 59 m 57 25 rf 4 232 From this statement it appears that, taking all four infirmities together, the average on the whole population begins to be exceeded after the 30th year amongst males, and five years later amongst the other sex, and the general liability to one or other of the four continues to rise from the earliest period. Insanity, in the case of males, reaches its highest intensity in the decade fending with the 45th year, where it amounts to 5 per 10,000. If the accidental variations at quinquennial periods ■ be excluded, this infirmity may be said to remain at this level till old age. Women show the highest ratio between 45 and 54, or 10 years later than the other sex. Deaf mutes of both sexes show that, if correctly returned, they last a considerable time, but the high ratio at the end of life is probably fictitiously swollen by the inclusion of those who are only deaf. In other respects, it is curious to see how the general average is maintained amongst both sexes at all periods except between 10 and 25, and at the extremities of life. Blindness begins to be markedly on the increase after 40 or a few years earlier, but takes its great leap after 55 amongst both sexes. Amongst the very young, though less prevalent than at any other age, it seems from the return to be more than 12 times as prevalent as insanity in the males, and 15 times amongst females. Leprosy, again, seems to pass by the young and to begin its attacks about . 25 years of age. Its maximum prevalence is at 50 years, after which those afflicted by it seem to die off. In the case of females, the maximum shown in the table is at 55, but this is a period which is safely taken as conjectural, and is distributable between 50, and 60, without much change of discrimination. The next series of figures that it is advisable to bring to notice here is that which shows the proportion of females to Table D. — Proportion of Females to 1,000 Males ^ at each Age. Age-period. Total Population. a es p Blind. Lepers. 0-4 ],039 651 728 651 582 5-9 939 614 633 633 685 10-14 799 637 553 638 542 16-19 943 646 603 695 483 20-24 1,079 601 621 737 412 25-29 1,004 510 604 809 339 30-34 972 555 614 928 315 35-39 875 578 569 952 262 40-44 950 628 632 1,054 273 45-49 S59 675 618 1,048 254 50-54 1,005 S24 674 1,179 276 55-59 922 729 657 1,125 .301 60 and over 1,199 967 826 1,125 361 Age not slated 909 850 709 1,23,1 507 All ages 965 637 630 990 321 of the passions ' 1,000 males at each age, in the four returns of infirmity respectivelyt The general proportion found to prevail amongst the population at large is added in Table D., for the sake of comparison, in anticipation of special notice in a future chapter. The main point is so prominent, in the figures as they stand that comment is almost superfluous. In no case but that of the blind does the ratio come anywhere near that found in the general return, and as regards the blind, too, the female element preponderates only after the 40th year. Xmongst the insane, at the prime of life, the women are but little above half the men in number, and this is in accordance with a tendency found to prevail not only in India, for the is more strongly demarcated in the sex which has given itself greater license to indulge them. It will be noted, too, that after the first period, which comprises, probably, the congenitally idiotic, the ages at which insanity in the females approaches nearest the proportion of xhis infirmity in the other sex, are those of early child-birth, and of the change .of life. After the latter period insanity tends to greater sexual equality. Amongst the deaf-mute, as has been already noticed, the small proportion of females is only what is found to be the characteristic of this infirmity in most countries where it prevails, and, indeed, of all congenita] defect. It is at its minimum at the age when there is the greatest inducement to coQceal it, which is also, it will be noted, that in which the proportion of girls generally is the smallest. From 40 years onwards the ratio begins to rise. The proportion of blind girls to boys similarly afflicted is curiously in consonance with that of the insane for the first three periods shown in the table. It rises, however, continuously from the fifth year, without intermission at the dangerous periods of life, and, as just observed, from 40 years onwards, the bliiid women are in excess of the blind men. The case of the lepers is very remarkable, so far' as the sex proportions are concerned. In the first place, there are three times the number of males that there are of the other sex. ' Then again, up to 30 years old this average ratio is exceeded, a fact which is not met with again until the end of life. In the first three periods slightly more than half the number of 233 lepers are females, but the ratio goes on decreasing till the 40th year, after which it oscillates irregularly, in liarmony with the general inaccuracy of the age returns. In treating of the territorial variation in the prevalence of the four infirmities Territorial in question, it is necessary to take into consideration the probable effect of relative distribution inaccuracy in the enumeration. To a certain extent this can be appreciated from the comparison of the two last returns, that of 1891 and that of the preceding census. The following statement will give some notion of the differences between the two, as Table B. — Showing the relative Order of the Provinces as regards prevalence amongst Males of Infirmities in 1881 and 1891. Province, &c. Insane. Deaf Mutes. Blind. Lepers. 1881. 1891. 1881. 1891. 1881. 1891. 1881. 1891. 1. Ajmer 3 11 7 14 3 5 15 15 2. Assam. 9 3 12 4 15 9 4 2 3. Bengal and States 5 5 1 2 11 11 2 3 4. Berar 10 14 5 15 4 2 1 1 5. Bombay and States 7 7 10 9 ■ S S 6 5 6. Sindh, &c. 1 1 3. 1 2 4 ^4 u 7. Lower Burma 2 2 11 13 10 14 3 4 8. Central Provinces 1'Z 12 9 10 7 7 5 6 9. Cborg 13 8 4 7 13 15 12 13 10. Madras and States 8 10 14 6 9 13 7 8 1 1. N.-W. Provinces and States 1S 13 8 5 5 3 8 7 12. Panjdb 4 6 2 3 1 1 9 11 13. Haidrabad 11 13 15 11 12 12 10 9 14. Baroda 6 4 6 12 6 6 11 10 15. Mysore 14 9 13 8 14 10 13 12 it shows the serial order of each province or State with reference to the respective prevalence therein of the four infirmities in the two years. It is not justifiable, of course, to set down the whole of the very large variation that appears to have taken place in some of the items to inaccuracy of enumeration or tabulation alone, as it is no doubt true than in certain parts of the country other causes have been operative, which account for a good deal of the discrepancy. In Madras, for instance, the recuperation of the population since the famine has added so largely to the supply of young children that, except in the case of congenital disease in the affected tracts, the preponderance of that class must have had some influence in bringing down the ratio of the infirm, and it is noticeable that as regards the only congenital disease with which the census has to deal, the variation in the results is least, on the whole, in the tracts where the infirmity was most prevalent in 1881. With the above qualification, the results given in the statement can be best judged by assigning to each item a "figure of merit" according to the stability of its place in the four columns. By this standard, Bengal, Bombay, and the Central Provinces are in a class by themselves, since in two cases they occupy the same relative position in the list at both enumerations, and in the two others they rise or fall by one place only. Below them we find Sindh, which is on a par with them in respect to two infirmities, but changes in the others to the extent of two places each. There follow seven items in which there was identity of position in one case only Of these the Panj^b comes first, as it has varied but by two places in a couple 'of columns, and by one in the last. The North-West Provinces come next by "one point. Burma and Haidrabad follow at the same mterval again. Baroda is worse by two more, as in one case it varies six places, and Coorg, where, as might be expected from the migratory character of a large portion of its inhabitants, there is no instance of identical position on the list, is but two beyond Baroda. There is then a considerable break in the graduation. Mysore and Madras show a variation of 15 places, as compared with the 11 of Coorg, Berar has 16, Ajmer 17, and Assam, of which mention in connection with this point was made above, has no less than 22 points of variance. On the whole, however, it may be said that the results show that, other things being equal, the enumeration was good m all the main provinces ; that the variation is probably real in Madras and Mysore, where the conditions are very different in 1891 from what they were 10 years previously, G if I 78338. 234 and that in Assam the preceding enumei'ation was admittedly deficient in accuracy. As regards Ajmer and Berar, it is uncertain which of the two returns is to be trusted. In the former case Mr. Bgerton writes, " In 1881 the results showed an unaccountable " increase on the figures of 1876, and the present record shows as remarkable a " decrease on those of 1881. Possibly both enumerations were inaccurate in opposite " extremes. . . . Compared with the figures of 1881 for other provinces, the " proportion of afflicted returned is below the average in about the same degree as " it was above it at the former census." Ajmer is undoubtedly subject io more chance of actual variation than most other parts of the country, as it lies embedded in foreign territory, is a place of passage between two or three large tracts on each side, and lies conveniently for expeditions to several centres of Brahmanic pilgrimage, which are always the favoured I'esort of the infirm of all religions, who lie in wait by the sacred pool without thought as to the troubling of the waters. It is worthy of note, too, that both here and in Berar instances are mentioned of the application by the enumerating agency of the qualification of congenitality to blindness and insanity, as well as to deaf -mutism. This misunderstanding tends, obviously, to the reduction of the number returned. In Berar, on the other side of the account, it has been found that in 1881 persons were returned as deaf and dumb who had become so in later life, so that the enumeration having been more careful on the occasion under review, the diminution is in the direction of greater accuracy. This part of the subject may be concluded with the remark that the statement we have been con- sidering shows that whilst in the case of all the other infirmities, a number of the items, varying from three, as regards the return of lepers, to six, in that of the blind, occupy the same position in the list that they did 10 years ago, the column for the deaf-mute return does not contain a single instance of identity in the respect, thus proving the difficulty of obtaining a consistent application of the definition of that infirmity. In considering Table F., opposite, in which is shown the prevalence of the four infirmities to the uniform base of 100,000 of the population, it is, perhaps, more interesting to take the infirmity rather than the territorial item, as the text for comment. In place, therefore, of considering the diflFusion of the four infirmities over any particular political unit, information on which point is available in the provincial volumes, the local prevalence of each infirmity will form the subject of the next few paragraphs. Insanity. To begin with insanity, it appears that at both enumerations Sindh and Lower Burma came first and second respectively in the order of prevalence. In the case of the former, the cause of its bad pre-eminence is not ascertainable from the returns, as they do not distinguish idiotcy from lunacy, so that we cannot say if the climate or water of the river, or the drought, or the indulgence in hemp smoking, or all of any of these are responsible for the results. In Burma, it is said, the people are excitable and self-indulgent ; and the returns show that the M6n, a more phlegmatic, or, perhaps, abstemious race, is comparatively free from taint. The population, too, as a whole, is better educated. There seem other considerations, moreover, to which weight must be attached. For instance, in Upper Burma the ratio is higher than in Lower, and ae there is not yet an asylum in the former, there was possibly less concealment of domestic cases from fear lest the sufiierer should be removed to a place of refuge by the State officials. Then, again, the high ratio amongst the Forest tribes in the Lower Province indicates the probability of much of the return being taken up by the cretinism that seems to follow mountain streams. There is one more fact connected with the Burma return that calls for remark, namely, that the woman adult is apparently more prone to mental derangement than the male. If this were due to ordinary physical change, we should find it more prevalent, but it is peculiar to the upper portion of this particular province, and is unexplained. Third on the list comes Assam, where, no doubt, the greater part of the malady returned is cretinism of the description just mentioned, which follows the main river. In 1881, Ajmer enjoyed this rank, but, as remarked above, its decadence is unexplained. We find the Panjab and Baroda have changed places, and the former is now sixth instead of fourth, and vice versa. The Baroda report is not yet to hand, and so comment on the figures is inadvisable, in the absence of the local experience of the Superintendent. The chapter on infirmities written by Mr. MacLagan in his Panjab report is well worth perusal, as he has treated the subject with care and in detail, reviewing the circumstances of his large and varied charge -from a high standpoint. It appears at least probable tliat most of the insanity in this Province is connected in some way with special rivers 235 Table F.— Showing in each Province, &c. the Average Number of Persons afflicted ;, per 10,0,000 of each Sex, in 188.1 and 1891. A. — Maies. B.--I)ba» Mutes. , Province, &o. ^ Males. Females. Males. Females. 1 , > ' 1881. iSgi. 1881. 1891. 1881. 1S91. il881. 1891. Ajiner 69 22 42 9 80 : 39 1 61 24 Assam 38 62 26 49 68 96 41 76 Bengal and States 53 4' 36 28 153 117 94 71 Berar 36 19 26 i 14 104 21 1 81 1 5 Bombay and States 41 3o 22 17 77 64 54 46 Sindh and Khairpiir 162 109 107 64 133 128 95 75 Burma, Lower 114 83 84 5i 72 42 48 34 „ Upper — 124 — 127 — 79 — 66 Central Provinces and States . 27 21 14 12 78 58 59 43 Coorg 23 26 18 25 108 80 85 63 Madras and States 38 25 28 18 62 83 50 62 1 North-West Provinces aind States 19 16 9 8 78 84 48 5o Panjab and States 59 38 36 22 141 iiS ii, 92 76 Haidrabad - • -i . ' 30 18 16 10 49 46 ' 29 3o Baroda . 51 43 34 27 93 45 62 3o Mysore 21i 25 14 19 68 77 56 61 Kajputana — i 32 — 19 — ■ — — — India 43 33 27 21 104 90 68 59 • C— I StlND. D. — Lepeks. Province, &c. Males. Females. Males. Females. 1881. 1891. 1881. 1891. 1881. 1891. 1881. 1891. Ajmer 355 181 588 208 9 6 3 3 Assam t 76 108 60 106 100 i83 40 61 Bengal and States 136 _ 100 144 96 122 93 41 3i Berar " i , 330 226 402 . 141 215 193 59 58 Bombay and States 217 140 273 149 • 84 77 31 26 Sindh and Khairpiir 391 209 561 221 13 7 10 6 Burma, Lower 152 89 162 99 101 92 33, 3i TJppei' — 317 — 416 — 160 — 81 Central Provinces and States 218 i56 296 188 89 66 41 33 Coorg 92 49 90 5o 25 12 23 14 Madras and States 153 97 170 99 68 54 25 18 North- West Provinces and States 270 216 290 224 63 54 16 i3 Panjab and States 488 338 538 368 61 29 21 II Haidrabad 128 100 110 84 42 38 19 i3 Baroda 248 161 351 235 39 32 17 i5 Mysore 89 106 98 104 16 21 9 II Bajpntana — 272 — 3go — 21 — 7 India 217 164 243 171 85 68 29 23 Grg2 236 such as the Ohenab in particular, along which cretinism is rife, from the Himalayan valleys to the plains of Multan!. This hypothesis receives some support from the experience given by the returns for Bengal, which province comes next on the list. Here, too, the Himalayan rivers seem to form the chief centres of mental unsoundness. On the other hand, so far as Multan is concerned, it is noted that the neighbourhood contains more than one shrine of repute for its curative virtues in such afflictions, so that the local idiots may well be reinforced by those brought to the locality by faithful relatives. There is also the fact that amongst a Musalman population, such as that of the "Western Panjab, there is greater latitude as regards consanguineous marriage than amongst the Brahmanic clans. But, taking the whole of the circumstances into consideration, there seems every reason to think that a specially baleful influence on the mind is exercised by mountains, whether the Himalaya or the Salt range, in Upper India. The detailed returns for the North- West Provinces have not yet been received in the Central Office, so there is a gap in this argument that has to be filled up by the local census Superintendent, whose province enjoys the honour of being, according to the results of his enumeration, the sanest in India. In Bengal, Mr. O'Donnell seems to think that the distribution of insanity is racial or professional rather that local, and that the Mongoloid strain in the eastern portion of the province, together with the wider expansion of instruction and the notoriously litigious temperament of the population in that tract, make, on the whole, a more favourable seed-bed for mental derangement than the stolid ploughman of Bihar, or the Dra vidian hill-man of Ohutia Nagpur. This may be so, but the details show that the cretin element is by no means unimportant, as the ratio of insanity is far above the provincial average in the sub -Himalayan districts and is remarkably high in the Tarai-bordering State of Kochh-Bihar. That the return, however, is not altogether free from suspicion is shown by Mr. O'Donnell's discovery of an attempt made by a righteous Hindu scribe to cast a stigma on the followers of the Prophet by entering all the infirm of his subdivision as Saiads by tribe. The errors due to fraud or otherwise seem, all the same, to have neutralised each other, for the province, like Sindh, Burma, Bombay, the Central and the North-West Provinces, takes the same place in the list as it did in 1881. As regards Bombay, there is nothing special in the distribution to call for remark, save that the general feature of the variation seems to show that enumeration on this occasion was more correct than that of 1881. It may be here noted that there certainly seems to be some connection between the returns of insanity and the general ccmdition of the people during the intercensal pei'iod, for wherfever the decade has been one of normal prosperity, the persons thus afflicted have fallen out of the census roll to a remarkably large extent, and this may be noted in Bombay, Bengal, the Pan-jab and Madras, as well as in smaller tracts. To a comparatively small extent, too, the spread of hospitals, and of dispensaries in the rural parts of the country muist have tended to check the growth of insanity, since the patient, who perhaps may be only temporarily deranged, dr epileptic, can be attended to by trained agency before the disease is confirmed. To return to the tables : — In Madras there are traces, according to the view of Mr. Stuart, the Superintendent, of a racial tendency, or predisposition to insanity, as the Uria and Tamil seem to be less subject to this mfirmity than the Telanga or the Malabari. The difiierence does not seem to be due to climatic or geographical causes alone, for the prevalence is by no means in general accord with that of the adjacent districts in the coast and Deccan portions of Bombay. There is one curious fact that may be mentioned for what it is worth, in the present state of the study of morbid pathology, and this is, that amongst the six highest ratios of the insane, no less than four appertain to communities of the Malabar coast, out of whom three are still polyandrous to some extent, in their domestic arrangements. Another fact brought out in the returns is that the ratio of insanity decreased duririg the decade to a greater extent in the worst of the famine districts than in any other portion of the Presidency. There is little that need be said regarding the rest of the provinces. The distribution of the insane in the Central Provinces seems to indicate that apart from cretinism, which is not prevalent there, there is nothing in the hill- country that specially breeds or wards off this infirmity. The Superintendent, indeed; mentions the tendency of the forest tribes to drink ' to excess, so that here is ready to hand the ♦"xplanation of the local prevalence of insanity amongst a class which elsewhere seems peculiarly though densely sane. jjgaf. The next part of the table to come under review is that which relates to the mutism f)roportion of the deaf-mute. Here we find Sindh. Bengal, and the Panjab again well to the front, both in 1881 and on the present 'OccasioiJ, Assam^ comes not far behind. 237 In all, except Sindh, where the source of the evil is not traced, it appears that the main prevalence is located in the HimEtlaya or alon^ the rivers flowing directly from that system, so that, as in the mountainous countries of Europe, this infirmity is combined with cretinism and goitre. It is possible that diet may have some effect on the system predisposing the inhabitants of the tracts in question to these forms of malady, for, at all events, in the Panj4b Himalaya the food of the people is very different from the staples in favour in the plains. This, however, does not explain the Sindh case, nor, probably, much of the prevalence in Bengal, Burma, or Assam. It is curious to find the same feature in the Burma returns from the hill tracts, whilst the M6n, as in the case of insanity, are less liable to deaf -mutism than the Burmese, and in Upper Burma, where there is a comparatively dry atmosphere, the ratio is far higher than in the lower division of the Province. Here, however, there is the probability of inaccuracy in the return, as the enumeration was conducted for the first time, and experience indicates that on such occasions the confusion of the deaf with those afflicted in both ways is greater than at subsequent inquiries. Still, the standard of life is lower, and the diet more sparse than in the Delta. In Madras we find the same feature as in Assam, namely, that the ratio of the deaf mutes to the total population has risen. This is attributed apparently, in part, as in the latter province, to bad enumeration in 1881. Beyond a general accordance with the distribution of insanity, and some slight preference for the hot atid dry districts of the South Deccan, the return for this Presidency presents no special feature. It is curious, all the same, to find that the increase in the ratio is least in the tract where the increase of population has been highest during the decade, as if the determining factor were the well-being of the masses. At the bottom of the list come Berar and Aimer, the two in which the variation is most clearly of all due to inaccuracy of return. In the return of blindness we find greater agreement than in the rest between Blindness, the results of the two enumerations, and the decrease, which is general throughout the returning tracts, with the exception of Mysore and Assam, where the ratio has risen, may accordingly be taken as truly representative of the facts. There are, as already noted, certain defects in enumeration, such as the inclination to enter, on the one hand, senile glaukoma and dullness of sight as blindness, and to include, on the other, those who were not born blind ; but it is not unreasonable to assume that these misconceptions are not universal, and, on the whole, are likely to be constant, and to balance each other. If the scale be unevenly weighted, it is probably on the side of excess ; that is, the undue entries are more numerous than the undue exclusion. On this hypothesis, the decrease in blindness in the last 10 years is in reality con- siderably greater than is shown in the returns, and the prpvincial Superintendents, as a rule, agree that this is the case in their respective charges. The improvement is attributed chiefly to two causes, first, the rapid spread of dispensaries in the smaller country towns, bringing trained assistance within the reach of many claesss who have been hitherto obliged to let ophthalmia and similar affections run their course amongst the children of the community. Then, again, though the increased accuracy of death registration somewhat obscures the fact on paper, there has been a notable diminution in the proportion of deaths to seizures in the case of small-pox, and this disease is, in India one of the most fertile sources of blindness. In Mr. Maclagan's review of the Paniib statistics these points are well illustrated, as he shows, as to the first, that the number of patients treated for eye affections at the public medical institutions of the province rose from 90,820 in 1881 to 233,670 in 1891 ; whilst in connection with the second t'lie comparative tables he quotes show that the decrease in blindness is greatest precisely in the parts of the Province where mortahty from small-pox has most declined and where vaccination has made most progress. This last fact proves how difficult it is to trace the relative prevalence of blindness m India to aily single cause ; for if locality alone were the determinant factor, the infirmity would be more diffused amongst the population of the glaring and arid plains of the Indus valley, instead nfrnachinff its maximum in the wooded and well-irrigated tracts ot the submontane - and easterl divisions of the province. On. the other hand, Sindh, hotter and dryer tlipn even the MuMn tract, is peculiarly favourable to the development of diseases of the eve but so appears to be the Malabar coast, where the climatic conditions are vaatlv different But, on the whole, the statistics for different parts of the same nrnvince— for the latter is far too varied in its component parts to be otherwise than i^n nnwieldv unit— seem to indicate that blindness is more prevalent, as a rule, m hot nd drv tracts and less prevalent in mountain air and within the ipfluonce of the heavier ' ■ ' Gg3 238 rain currents. In the Panjdb, for instance, th,e Himalayan tracts are markedly favoured as regards eyesightj so are the Chiitia Nagpiir and, other hill tracts pf Bengal. Along the coast, again, Orissa, the Konkan and the littoral tracts of Buripa present a marked contrast in this respect to the plains of Bihar, the Nortl)., Deccan, Upper Burma, Rajputana, and the adjacent and somewhat similarly conditioned tract of Grujarath. There are, however, the anomalous cases of the Malabar coast and the north-eastern coast tract of Madras to be considered. The statistics of the rest of the Presidency are in general accord with those for similar tracts elsewhere, that is, the Mlgiris and surrounding hilly tracts show least blindness, and the hot and dry plains of the Ceded Districts return the maximum. Mr. Stuart has accordingly subdivided the coast districts into; their littoral and inland tracts, with the result that the former seem to show more blindness than the latter in some cases, though not in all. It is possible that the geographical peculiarities of these tracts may have some influence on the variation, as there is a great difference in the distribution of the rainfall between Malabar, where the Ghdts lie at some distance from the shore, and Kanara, where they come close down to it. In the latter, the coast receives the full current ; in the former, it is attracted to some extent, inland. Perhaps there is some feature of this sort in Vizagapatam and its neighbourhood on the opposite coast, but in either case the suggestion is purely conjectural. It is also open to question whether, so far as ophthalmia and cognate disease is concerned, especially amongst infants, the prevalence does not to some extent depend upon what we may call normal agricultural dirt, which in India is often associated with high cultivation. The better the tillage, the nearer is the manure to the master's eye, and the dryer the air, the more numerous and busy' are the flies round every source of moist nourishment, from the dunghill to the eye of the sleeping infant. Pinally, and as regards the prevalence of blindness amongst women alone, there are the social considerations, such as seclusion within dark rooms' often reeking with acrid smoke from the cooking-places, which is customary rather in Upper India, Sindh, the Panjab, Rajputana, and the JSTorth-West Provinces, than in the rest of the country ; also the conventional mourning of relatives, which is necessarily very frequent, and is invariably accompanied, as Mr. Maclagan puts it, by much ostentatious squeezing of the eye. Whether either of these two causes is in operation on the Malabar coast is not stated in the Madras report, though no doubt the long rainy season keeps both sexes indoors to a greater: extent than in the light fall of the uplands, and may thus serve the same purpose in connection with 'our present subject, as the zenana system of the north. Reverting to the consideration of Table P., it should be remarked that in 1881 the ratio of the blind was higher amongst females than males in all parts of In^ia, except Coorg, Assam, and Haidrabdd, In the last two the peculiarity has been maintained on the present occasion, and extends to Mysore and BengaL Looking at this fact, and the tendency in the same direction in Madras, and in the opposite one in the case of Sindh, the Panjab, North-West Provinces, Ajmer, and Rajputana, it seems that where the sexes are nearest numerical equality, or the women exceed the men, blind- ness inclines to be less markedly a female affection ; and when the males are far in excess, the women are more prone to blindness. It has not been suggested, however, in any provincial report, that the objection on the part of the patriarch to the return of the females of his household at the census does not extend beyond those who are sound in mind and body, which seems a not unnatural inference to be drawn from the above figures. The relative accuracy of the return for the two years has been discussed on general lines above. All that it seems necessary to add is, that apparently Ajmer and Berar, possibly Sindh and Mysore, and admittedly Assam, show differences which can only be ascribed to the enumeration, without basis in fact. It will be noted that in both years the Panjab heads the list, though with a remarkable decrease in the prevalence of the infirmity. Berar and Sindh have exchanged places as second and fourth respectively, not without suspicions of inaccuracy in both cases, as the ratio in Sindh seems to have receded by one half. A precisely similar movement has been executed by Ajmer and the North-West Provinces for the third and fifth places, in both instances with a diminution in prevalence amounting to over the half. In the last-named tract the improvement is said to be real, and due to the extension of vaccination and the erection of rural dispensaries. The three next Baroda, the Central Provinces and Bombay — occupy the same position respectively as in 188] , a fact which shows fairly uniform accuracy of return. The decrease in ratio in all three is less, however, than in the tracts where blindness is more prevalent. 239 Bengal occupies the eleventh place in both years, and Haidrabad the twelfth. Prom the Central Provinces downwards, in the case of the males, and from Bombay in that of the other sex, the ratio falls below the general mean for India, showing how much more this infirmity prevailsin the north than in the centre, south and east. EajputSna and Upper Burma come on to the list for the first time, and, as in respect to other infirmities, no doubt show a fuller return than will be the case at future enumerations. Both, however, are hot and dry in climate, so that the ratio, though likely to decrease, is not to be expected to descend to the level of that of coast tracts. Summarising the results of the two returns, it appears that blindness tends to be more prevalent in hot and dry plains away from the hills and sea ; that in the north df India social customs are specially favourable to its spread amongst women, but that whether the infirmity be due to congenital or infantile ophthalmia or to small-pox, it is everywhere on the decrease owing to greater facilities for obtaining timely surgical _ assistance, and _ to the diihinution of smali-pox due to the extension of vaccination. :' ; The subject of leprosy has been brought very prominently before the public Leprosy, during the last three or four years, and so far as India is concerned, has been exhaustively treated by a special commission of qualified experts who have analysed all available statistics in the report recently published.* The record of the results of their investigations, therefore, will necessarily supersede anything that may be written on the subject b;^ a layman, so that a short shrift and sharp execution is all that the statistics require in the present work. Table E. shows that in only three cases do the positions of the territorial items correspond in the two years. Bei;ar still heads the list, and the small variation in its ratio, both amcmgst males and females sho'w that if inaccuracy there be, it is constant, and therefore statistically insignificant. Sindh and Ajmer again occupy the two last places, each with a considerable proportional decrease in the prevalence of their lepers. As regards the rest, it will be noted that none but Assam has altered its position by more than a single place, and the explanation as tp the variation in this last has been already given. It is doubtful how far locality is connected with the prevalence of leprosy. From the fact that the ratio is high in the hill tracts of the Panjab, and in the corresponding portions of the North- West Provinces and Bengal, and on the Toma tracts of Burma, it appears that mountainous country is favourable to its development and a dry if not hot, atmosphere inimical. There is some evidence, too, of its affinity to coast tracts ; but here, again, the Malabar figures are inconclusive, for those of South Kauara contradict their companions of the next district. In another direction, it appears that this disease finds 'a poor and ill-nourished population more accessible to its inception than one whose average condition is higher ; and it is also stated that there seems to be some connection between the circumstances giving rise to cholera and those favourable to the development of leprosy. The Report of the Commission just mentioned, of which a summary is all that is at present available for the purpose of the present work, is decidedly encouraging in some of its conclusions, and, speaking generally, there seems reason for anticipating that most of the influences which predispose the population to the attacks of this disease are such as are likely to give way as the standard of maintenance advances. There is only one further point in connection with this subject with which it is infirmities necessary to deal here. It is the ; respective affinity between the four infirmities by caste selected for. investigation and social circumstances, so far as the latter can be judged S^oups, &<=■ by, caste or race. We must first, however, revert for a space to what was said above as to the influence exercised on certain forms of disease by climate and locality. These, it will have been seen, affect, to a greater or less extent, all four of the selected infirmities, so that, looking at the varied territory comprised within the limits of a single Province or large State, it is most probable that distinctions of caste in respect to s^ch matters will, he outweighed by physical conditions. As a strildng instance of this we may take the prevalence of deaf mutes amongst the two widely difiereut oasfces of the Brahman and the Ohana^r, or leather worker in the Panjab- The ratio is about the same 'in both, but more than a third of the total number of affected persons are returned from a single district in. the hill tract. But if we compare the proportions amongst the different caste groups for all India, we obliterate. * A copy of this, Report only reached me on the 14th June 1893, so the contents could not be fully adopted in thjs chapter. — J. A. B. ,„ . ' T^'-- ■ ■ - (. g 4 240 Castes showing a generally kigh ratio of infirm. m turn, local influence, and any distinctions there may be are most probably inherent in the social, not the geographical or climatic circumstances in which the caste in question is situated. The figures on which the following remarks are based will be found in the different sections of Table XVIII. in Volume II. of the Imperial returns. It was necessary to split up this table into three, because the same detail was not tabulated in every one of the constituent tracts. Rajputana, for example, tabulated three of the infirmities by caste, but not deaf-mutism, whilst Central India and Kashmer omitted this portion of the schedule for all but the cantonment population, &c. It is not proposed to enter into territorial detail in respect to the results of the tabulation, for the reason just given above, nor does it seem worth while to reproduce the whole table in its proportional form, as its beariftgs can be adequately exhibited by the following extracts. Table G-.. below, for instance, gives the 12 caste groups in which the total number of afflicted is highest relatively to the total strength of each sex in the group. It should be explained that betel leaf selleTs and brass-smiths are amongst the 10 highest in the case of males only, and their place amongst those of the other sex is taken by the leather- working castes and the barbers. The first Table G. Males. Females. Caste Group, &c. CO s to 1 4J 1 ffi 6 1 S i C3 .= t-t 3 i "? C3 "a i 0. 3 fl ^ ^ OD 00 1 00 'S 5, 00 00 i-< fco Under 1 1,012 1,006 9S0 1,002 1,015 1,072 1,057 998 962 952 956 975 1,001 990 Total under 5 986 977 1,036 988 1,060 1,076 1,054 1,003 976 971 964 976 1,011 1,000 5-9 897 876 915 895 911 959 1,027 1,006 976 977 966 978 1,004 1,003 10-14 765 752 7ZI .733 781 796 898 997 968 957 959 980 1,011 1,012 15-19 816 783 822 876 828 1,038 905 1,008 994 1,040 1,025 990 1,043 1,124 20-24 965 964 1,016 1,033 1,002 1,174 1,098 1,093 1,063 1,060 1,017 1,034 1,046 1,173 25-29 847 863 967 965 921 910 994 1,087 1,105 1,120 1,027 1,066 1,052 1,007 30-34 860 903 9C0 921 912 983 986 1,077 1,117 1,161 1,027 1.097 1,067 1,045 35-39 805 836 852 810 851 762 867 1,069 1,161 1,128 1,015 1,113 1,055 957 40-44 915 949 882 88S 964 917 907 1,079 1,166 1,128 1,010 1,118 1,050 1,000 45-49 691 811 788 715 852 706 852 1,103 1,204 1,047 996 1,113 1,072 95Q 50-54 945 952 960 763 957 961 1,050 1,104 1,215 1,095 1,020 1,134 1,146 1,097 55-59 706 824 890 683 834 749 960 1,111 1,224 1,058 986 1,144 1,104 1,001 60 imd over 1,085 1,146 1,190 766 1,156 1,158 1,246 1,187 1,377 1,075 980 1,289 1,068 1,022 Total 881 891 913 882 928 964 991 1,055 1,076 1,043 995 1,061 1,047 1,030 or that which follows it. Both these causes tend to reduce the proportion of the women in the period in question, and there is also to be taken into consideration the probability mentioned above, of the omission altogether of a girl of this age who has not been provided with a husband, or who, in Upper India, is attached in some intimate capacity to the zanana of the richer classes. But in addition to these artificial reductions of the number of the females of this age, it is not at all improbable that there is a real deficiency, due to the fact that amongst the great majority of the population the five or six years in question include the first child-bed, an occurrence notoriously dangerous to female life, especially where the wife is as immature, physically, as she too often is in India, and vrhere the obstetric methods in vogue there tend to restrict survival to the fittest only. It must be borne in mind that at this time of life np less than 83 SO per cent, of the women are manned, whilst only 36''80 of the males are in the same condition, and in spite of the general deficiency of the former sex, which here amounts to over 850,000, the wives are more than twice the number of the husbands. Carrying this train of inquiry back to the preceding period, where the disproportion is, as observed above, at its height, we have included but a year, or, at most, two, of possible child-bearing, but the arrival of the girl-bride at puberty, an event that is almost invariably accompanied with certain •ceremonial rites of a somewhat exciting character, and which entails also other demands on the constitution, which are not, in the circumstances, duly recognised, must have its weight in considering this question. The discussion that arose in the course of last year's legislation regarding what is commonly known as the " age of consent," amply shows that the practice of premature cohabitation is more or less local, or restricted to certain castes, as is indicated by Mr. O'Donnell in his report on the Bengal returns of marriage, but none the less is this period a critical one for girls in India, if only on the ground of the demand on the nervous system, for we find that irrespective of the second and later part of the conj ugal arrangement in that country, out of the 13 millions or so girls between 10 and 15 years of age, 49^ per cent, are married in the Indian acceptation of that term, and nearly 1-^ per cent, are widows. The conditions under which the other sex has to struggle onwards at this time (if life are far less importunate on the constitution in this respect, for only 15^ per cent, are husbands, and less than ^ per cent, widowers. Entirely apart, 249 therefore, from any question of concealment or other cause of omission, and of errors in statement of age, so extensive and remarkable a divergence of practice betvr een tlie two sexes in reference to such an important and universal factor in social circumstances as marriage is not at all likely to be without its influence on the distribution of the sexes at the periods under discussion, and this influence, if admitted, must inevitably be cast on the side of the results indicated in the returns of both the last enumerations.* It is not worth while to pursue the analysis of the return in detail, because the preference shown, especially by the fair sex, for the even multiples of five, which has been mentioned already as a grave defect in the age tables, grows more and more apparent with age, and even if we adopt the simple method of combining two periods, as from 25 to 35, and so on, the curve is still "backed like a camel" from the age of 24 downwards. The general tendency, however, in India, as in many other countries, is for the women who have safely weathered the period of early childbirth to outlive the men, but there are exceptione to this rule, as in Sindh, Kashmir, the Panjab, and, curiously enough, Lower Burma. Then, too, there is no doubt but that the age of the old women are exaggerated, and that, though their life is certainly somewhat better than that of men in advanced years, the difference is not so great as that indicated by the returns. It is obvious, again, that so far as the masses are concerned, the wear and tear of woman's life is much the same in India as that of the man's. She shares in most agricultural operations, barring those in which the plough is concerned, carries the grass and. (irewood to market, and is largely concerned in all general labour, particularly of the class in which the outside of the head bears the brunt of the toil. There is a strong probability, accordingly, that the deficiency shown irregularly between the ages of 35 and 55, has to some extent a foundation in fact. Amongst certain classes, too, as Mr. O.'Donnell points out, remarkable prolificity in the prime of life tends to wear out the the woman at a comparatively early age, and the distribution of the Musalman population of Eastern Bengal is cited as a good instance of this. In the table on page 179 of the Bengal Keport the statistics in question are shown side by side for Hindu and Musalman. The proportions are slightly in favour of the latter between 10 and 30 years of age, after which period the decline in the case of the convert is remarkably swift, and after 60 years of age, there are but 951 Musalman women to 1,000 males of that faith, whereas the Brahmanic classes number 1,179. The general tendency amongst forest tribes seems to be in the same direction. The women are' on an equality at least, numerically, with the men, up to a certain age period which oscillates between 85 and 45, after which the hard work of semi-savage existence, rather than the pain and peril of childbirth, has its revenge. In nearly every province, too, the age tables, even in their uncorrected state, show that amongst Indian women, as elsewhere, but to a more marked extent, the time at which the change of life takes place, which is generally between 40 and 45, is a very critical one. The above resume of some of the more obvious factors afiecting the distribution Climate and of sex in India does not, it may be noticed, take into consideration the influence Nutrition as of two elements to which there is no doubt but that some weight should be attached. j° g"^"*^^^ The difficulty with regard to them is that they are not mutually exclusive, and the distribution, sphere of their influence is, at best, but vaguely definable from the available data. The factors in question are, of course, climate and nutrition. A review of the whole field of statistics resulting from the census inquiries seems to afford ground for the following deductions, which, however, are not put forward for the present as more than conjectural. The ratio of females to males, taking the whole population in existence at one time, has a tendency to be higher along the coast or within the influence of sea air, to an extent beyond what can be accounted for merely by the temporary absence of a certain number of males at sea. It runs higher, too, in hilly tracts, as a rule,, than on the plains, and it seems to be depressed by a dry and hot climate, particularly if accompanied by a considerable range of temperature. On the other hand, wo find traces of the influence of nutrition, which in some cases -may fairly be held to neutralise that of climate. It is difficult to prove beyond a doubt any of the above tendencies, for the reason given m the beginning of this chapter, that in the determination of sex so many factors probably * It is not improbable, too, that the prevalent ciistora as regards dress may be prejudicial to female vJ^.lU,. hPfween 5 years old and 10 or U; since, as a rule, when children take to clotlies at all, the boy wears only a jacket, but thus covers all above the waist; but the girl wears only trousers or a petticoat, and leaves the more vital organs unprotected.— J. A. B. / 78388. I i 250 enter that, in the present state of information, the relative influenqQ,of each cannot be accurately discriminated. Then, again, m a country like India there must be anomalous cases in every such conjecture, which obscure the view of the operation of th,e; general rule, if one there be. A few instances of this may as well be cited Here, to show the difficulty of dealing adequately with the subject. The coast tracts are, as a rule, highly fertile, so that the general predominance of females in their population is attributable to good nutrition as well as to climate or situation. But where we get a comparatively ill-nourished population in this tract, as in the case of .Eatnagiri, on the west coast, and thus eliminate one unknown qua,ntity, we only find that a second is to be tackled, in the shape of the emigration of males to other, scenes of labour. In the Panjab, again, there is a hot and dry tract lying adjacent to Sindh, but the ratio of women is not at its lowest there, but in a fertile and highly cultivated triangle further north, the population of which are suspected of unfair treatment of the weaker sex. Then in Sindh the ratio is at its lowest, and the country is hot and dry, but it cannot be said that there is any failure in the food supply of the province, so unless we accept inaccurate enumeration as responsible for the whole of the inequality, there is clearly some important factor not as yet ascertained. In Rajputana, where the improvement in accuracy between the two enumerations is more marked than in any other part of India, the lowest ratio was not in the desert, but in the fertile plains bordering on the Jamna, where the existence is suspected of causes probably not dissimilar to those that prevail in the Central Panjab. The inconsistency between Bihar and the upper portion of the Grangetic basin in this respect, has been already brought to notice. Then as regards Bengal, where there is no sterile and no dry tract, for the Chiitia N'agpur plateau falls within the belt of light but certain rainfall, further complications enter the problem. Admitting, with Mr. O'Donnell, that the zone of deficiency of females corresponds with that occupied by races of Mongoloid parentage, we have to ascertain whether the difference is inherent in race or attributable to artificial causes, such as that of extreme fecundity and consequent high mortality amongst women at a comparatively early age, for climatic considerations have clearly nothing to do with the case, and those of nutrition tell in a direction opposed to the facts, since, so far as Eastern Bengal is concerned, the standard of living is above that of South Bihar. The tide of immigration in this province sets from west to south-east, so the birthplace returns should throw some light on the conditions. On eliminating the immigrants, it seems that in both Northern and Eastern Bengal the ratio rises from 966 to 996 and 989 respectively, but is still far from the corresponding figure for the rest of the province. But in Assam, again, we find the same deficiency of females as in Eastern Bengal, for, deducting the immigrant population, the ratio is only 969 of that sex to 1,000 of the other, and here, too, the deficiency is in the fertile valleys, not amongst the hill population. This seems to point to some conditions common to this province and its neighbour, probably social or physical rather than racial. In Madras, the difPerence between the Deccan population and that of the south and west as regards the proportion of females is very marked. Mr. Stuart has shown that it extends to all castes, and points oat how a Tamil caste, such as that of the traders (Ohetti), returns a considerable excess of females whilst the class exactly corresponding to it in Telingana (Kdmti)'is in the reverse position. It is the same with both higher and lower castes than this, but it does not seem necessary to invoke racial characteristics to explain the discrepancy, as there is strong differentiation in both climate and nutrition which may account for it. There are two curious facts noted by Mr. Stuart in dealing with this interesting subject. First, that it is the general belief amongst the people of the southern tracts of the Presidency that 30 or 60 years ago there were fewer women than there are now, and that whereas formerly wives were at a premium, in the present day they have to be accompanied by a dower. The returns show that it is true that in those parts the proportion of unmarried women of 15 years of age and upwards is higher than it is in the Deccan districts. Then, again, so far back as the beginning of the century, no less competent an observer than Sir Thomas Munro found ample reason for agreeing with the popular belief in the " Ceded Districts " that there were only nine women tq every 10 men. He tested the accuracy of the statement by personal investigation in different parts of his charge, and satisfied himself that, taking the whole of the tract together, the men were in excess of the other sex. Assuming this fact, the higher proportion of women now returned seems to argue in favour of the nutrition theory. We are met, however by a case supporting the opposite conclusion in the Gujarath division of Bombay, where the ratio of females is lower than in the Deccan, in spite of the general fertility of the tract and its accessibility to sea air. It is noticeable, however, that the ratio tends to 261 descend in a northward course, and as in that direction the tract is bounded by Sindh and Eajputana,, where the great deficiency of the sex cannot be wholly attributed to anything but inaccuracy qf return, it is probable that a good deal of the Gujarath discrepancy also can be accounted for in that way. In Burma there is a great difference between the two divisions as regards sex distribution, as. in the upper province females are largely in excess, whilst in the lower they number no more than 892 to 1,000 males. It appears, however, that the latter ratio is due to immigration and temporary sojourners, amongst whom the ratio of females is 694 in the case of those from Upper Burma, and only 230 amongst the Indian- born. The birthplace tables show that the home-born have a preponderance of females, as in the upper province, and that along the coast the ratio is generally higher than inland. In no case, however, does it reach that of the newly-acquire.d territory, of which a few frontier districts in the north and east, where no doubt there is an influx of temporary visitors from the Shan hills, are the only ones that do not return a considerable excess of females. To some extent this is accounted for by the fact that whilst this tract was under Burmese sovereignty, emigrants to British territory were forbidden to take their women with them, and since annexation, there being no restriction, the proportion of females to 1,000 males in that class has risen from 598 to 694. This movement does not, however, suffice to wipe out the distinction between the two divisions of the province, so that, on the whole, the case tells slightly adversely to both the climatic and nutrition hypotheses. To summarise what has been said above, it seems that in most parts of India Summary, proper there is a tendency, in a greater or less degree, to omit from the census record girls of from 9 to 15, and wives of from 15 to 20, or thereabouts, but that in every part of the country, except the north, girls below five years old were returned as more numerous than boys of that age.- After that period, apart from wilful or ignorant omission, there is probably a real deficiency in the number of females, extending to about the twentieth year, more or lessj and due to neglect, functional excitement, premature cohabitation, and unskilful widwifery. At a later period, hard work, as well as the results of the above influences, and, amongst some classes, excessive fecundity, tell on the female constitution, producing greater relative mortality than prevails in the other sex, though towards the end of life the latter succumb to old age sooner than the survivors from amongst their mates. It is also probable that either from dift'erence or inferiority in nutrition, or from climatic influences, female life is on the whole, better in India on the coast and hills than on the hot and dry plains. So far we have treated of the circumstances that affect the distribution ^of the Sex at sexes amongst those actually in existence. There remains the question of whether birth, the census throws any light on the determination of sex during the " nine months' ante-natal gloom."* Here we tread on delicate ground. In one of the more recent works on the subject the authors say : — " The number of speculations as to the nature of sex has been well nigh doubled since Drelincourt in the last century brought together 262 'groundless hypotheses,' and since Blumenbach quaintly remarked that nothing was more certain than that Drelincourt's own theory formed the 263rd. Subse- quent writers have, of course, long ago added Blumen bach's Bildungstrieb to the hst ; nor is it claimed that the generalisation we have in our turn offered has yet received ' final form,' if that phrase, indeed, be ever permissible in an evolving science, except when applied to what is altogether extinct." The hypothesis in question, as stated by its authors, is that — " Such conditions as deficient or abnormal food, high temperature, deficient light, moisture, and the like are obviously such as would tend to induce a preponderance of waste over repaiv,— a, Katabolic habit of body —and these conditions tend to result in the production of males. Similarly, the opposed set of factors such as abundant and rich nutrition, abundant light and moisture, favour constructive nrocesses i e ' make for an Anabolic habit, and these conditions result in the production of females. With some element of uncertainty, we may also include the influence of the age and physiological prime of either sex, and of the period of fertilisation." Now in India, as in Europe, the male births everywhere exceed the female, and, looking at the diversity of ratio between the two in countries where registration is fairly accurate and complete, there is no reason to suppose that in the tracts we are now considering, there is more than a slight degree of incompleteness m the data for females as compared with those for the other sex. That there is some failure m this resnect is admissible, though, as some of the Superintendents have pointed out, the information is not furnished by the parents of the newly-born infant m rural tracts, * Geddes' and Thompson's " Evolution of Sex." 11 2 252 but obtained indirectly by the accountant, who knows every house in his charge, and has no more interest in returning boys than girls. Where the vicissitudes of season are very considerable, one must expect to find great irregularities in the birth ratio, and in the case of a large proportion of the masses the average standard of nutrition is decidedly low. The best examples of birth registration are given in the Madras Census Eeport, and, amongst other noteworthy features, they exhibit the peculiarity of containing the highest ratio of male births in the tracts where the general proportion of that sex is the lowest, and vice versa. The confirmation these returns lend to the theory above quoted is very slight, still there does appear to be in them some ground for holding that male births have the tendency to increase relatively to those of females as the amount of nutrition gets lower, and to decrease, conversely, as times improve. There is, again, the question of how far the respective ages of the parents affect the sex of their offspring. The evidence collected by statisticians in Europe on this point begins with the results of the inquiries by Sadler and Hofacker, who held that the greater the distance between the age of the parents, the higher the chance of the offspring being male. Since 1830, however, investigation on this line has been extended over wider areas and larger collections of people, with the result of severely shaking our faith in this theory. At all events, the statistics with which we have to deal at present do not seem to give fruitful results. Finally, there is the supposed tendency of inherited volition. Where a population has special reasons to desire male or to deprecate female children, it has been said by some that a high preponderance of male births will probably ensue. But such an influence can only be a feeble one, for the reason that its extension would in time tend apparently to the extinction of the race ! In conclusion, the information on sex distribution in India furnished by the census inquiries indicates that, as in most other countries, more boys are born than girls, but owing to the very much higher mortality amongst the former during the first vear of life, the latter predominate in number until their vitality begins to be affected by special sexual influences from which the male is free, so that throughout almost the whole of the prime of life the females are to a greater or less degree in defect, and the balance swings back only towards the end of life, when the total number is insufficient to restore numerical equilibrium. CHAPTER IX. The Population by Civil Condition. " For what secures the civil life But pawns of children and a wife ? That lie like hostages at stake, To pay for all men undertake ; To whom it is as necessary As to be born and breathe, to marry. So universal all mankind In nothing else is of one mind ; For in what stupid age or nation Was marriage ever out of fashion ? Unless ." — Butler. As in the case of languages to the philologist, so as regards the study of the branch of ethnology that concerns itself with marriage custom, there is no field more varied and worthy of research than that presented by the population of India. It is but a minute corner of that field, however, that lies within the sphere of exploration to which the census is restricted. A return of the people by civil condition was first required at the enumeration of 1881, and the inquiry was followed up on as far as possible identical lines in 1891. The instructions for filling up this column in the schedule ran as follows : — 1881. Young boys and girls who may have been married should be entered as married, even though they may not have actually begun to live with their wives or husbands. A male or female whose first wife or husband has died should be entered as widower or widow, unless he or she has married again, in which case he or she should be entered as married. 1891. Enter each person, whether infant, child, or grown up, as either married, unmarried, or widowed. This column must not be loft blank for anyone, of whatever age. Children who have been married should be entered as married, even though they may not have actually begun to live with their wives or husbands. Persons who have been married, but have no wife or husband living, should be entered as widowed.. The enumerator must accept the statement made by the person, or, in the case of children, by their relatives. The inquiry was thus limited to the simple question of whether the individual was single, married, or widowed. It was above all things necessary to avoid the appearance of any undue curiosity on the part of the State as to domestic details, hence the prohibition at the end of the rules of any controversy on the subject on the part of the enumerator who had to fill up the schedule. The statement of the person concerned was to be taken as final. There are certain special difficulties, however, in obtaining even the apparently simple information sought for under the above instructions. Several of the census Superintendents note one or two points in which there has been confusion of condition. For instance, in Assam, though there are plenty of words for widow, and a special word for bachelor, none was found, in the Brahmaputra valley at least, for widower, so the two classes of men who had not a wife living, whether they had had one already or not, were sometimes entered under one head, 'and, curiously enough, that of widower. Where, too, there are special colloquial terms for the widow who has married again, it vras sometimes held that the general word ought not to be used, so the second ceremony was ignored, and the wife entered as a widow, or, in Assam, where the case is particularly prominent, under the special title. On the other hand, some men who had lost their wives, are said to have been returned as married, especially if still in the prime of life. The question, again of how to enter the divorced arose in some parts of the country. In Burma, where the institution is most in vigour, there is a recognised term, which was returned without, hesitation. In India proper, though in every form of creed prevailing there the separation of husband and wife, under certain formalities, is condoned or permitted, the status is by no means recognised by either sex. It was the intention of the rule above cited that such persons should be entered amongst the widowed, as they had been married, but at the time of the census had no partner living. The distinction, however was too subtle for the commonalty, and in many parts of India, even those li 3 Polygyny. 254 wlieve divorce is not rare amongst the lower classes, the former stakes was returned, and both parties appeared amongst the married. This is not the case, it seems, in some of the hill tracts, where the number of widowers and widows is clearly due to the freedom of closing the conjugal union qtherwise than by the strong measure of leaving the world. Finally, in almost every section of the lower classes, particularly where the caste rules regarding marriage expenses are stringent, there are numbers of cases in which the wife is one in fact only, but not united by any of the bonds of ceremonial which play so leading a part in the system- at large. This practice of permanent concubinage is necessarily confined to those whose property is either insignificant or of a nature not likely to form the subject of a disputed succession. The woman is returned at the census as married, and as often as not the man who entertains her has lost his legitimate wife, and has thus every title, according to his standpoint, to enrol himself in the same category. We have also to deal with the converse case, of the wife who has, and never had, any husband, but who was formally wedded to a dagger, a fig-tree, or bunch of flowers, in order to give her the more dignified status. The above constitute the main points with regard to which the statistics now to be reviewed are to some extent untrustworthy, but, speaking generally, the distribution of the population under the three heads of married, single, and widowed, can be accurately gauged by them. It should be mentioned here that Table VIH'. ^ Volume'!, of the Imperial series excludes the population for which these details are not available by reason of the omission from the schedule of the column referring to civil condition. The balance, however, ambunts to over 262^ millions, the largest population for which this class of statistic has hitherto been collected. : But though the amount of information thus obtained serves the purpose for which it was intended in connection with the census, especially- as it has been tabulated in combination with age, and in some Provinces and States with caste also, its compre- hensiveness tends to level the diversity and to obliterate most of the distinctions in which the ethnologist delights to revel. ISTot to go beyond very general topics, we find in the returns comparatively little that is new on the subject of polygyny, save that it is alraost entirely restricted to the richer members of the Musalman -community, on the one hand, and to sundry hill tribes of the Central Belt and the north-east Frontier on the other. The extent to which it exists amongst the Brahmanic sections of the people must be very slight, though we are aware of the fact of its legality, and know that cases are not rare in which , a second wife has had to be introduced into the family circle, owing to the failure of an heir by the first. Taking the whole of India, without Burma, there are seven in every thousand Benedicts who have more than one wife. The provincial statistics are liable to be disturbed by migration of married men for the season, as well as by the custom, in so many parts of the country, if not universally, for the wife to return to her father's house ' foj her first childbed, as well as by the custom of dancing girls, &c. to return themselves as married, though their spouse was a dagger or graven inaage. Balancing the omission of young wives mentioned in the last chapter against the widows and others', shown as married, though livmg in concubinage, the excess of wives over husband^ seems to be greatest in the south of India, where it is estimated to amount to nearly 40 per 1,000. But here, too, there is a general excess of women, and it is possible that the census of this sex was more accurate than in neighbouring tracts. In Upper India, where the proportion of wives to husbands is in remarkable contrast to that of women to men a,s a whole, there is no doubt polygyny to a small extent amongst the richer classes, as stated above.' In the Central Provinces, where the sexes are generally much, op. an equality in point of number, there is a remarkable difierence between the forest tribes and the rest of the community 'in, regard to the excesS of wives over husbands, showing the prevalence of polygyny' amongst the former to the extent of more than even the Madras ratio, and toucliing, in some cases, 60 in the 1,000. In the neighbouring province the same feature is found. Polyandry. In the opposite direction, too, India has its examples to bring forward of polyandry, or the plurality of husbands. In its patriarchal and simplest form we find it in the Himalayan valleys, chiefly wherever the food supply of the surrounding country is scarce. It is found in the north-west amongst. both. Buddhists and Brahmanic tribes, so it is not attributable, in these tracts, to religion, though we find it cropping up again at the eastern end of the Himalaya amongst the latter; But in the Panjab, certainly, if not elsewhere, the custom of the marriage of onejwoman to the whole family of brothers is on the wane, and is growing to be- regarded as disreputable. Peace and plenty may have as much to do vrith this change of sentiment 255 as convictioB. In the northern plains of India, again, there are traces of this custom in certain tribes, and even in the south of the i)eccan, Mr. Stuart has heard of its existence amongst a subdivision of the great Reddi caste. But the great stronghold of polyandry in Southern India is on the Malabar coast, where it prevails amongst many of the more numerous castes of different social positions. It probably originated, SQ far as these castes are concerned, -with the Nair, or ruling community on that coast, who follow the matriarchal form, not that of the Himalaya. Here the succession is through the female, and the rules thereof somewhat complicated. In fact, the census officer for the Travancore State considered that the term married was not of itself sufficient to denote that condition in a population under the " Marumakkathayam " system, as the husband by marriage is not the husband by consummation, nor is the wife of the male member of the community the woman to whom he has been united by marriage rite. It is proposed tha,t to meet local peculiarities in this respect the present column of the schedule should be expanded to eightj in the case of males, and nine in that of the other and favoured sex. As a third instance of the special factors that enter into the consideration of Levirate. Indian matrimony, we may mention the practice of marrying the widows of near relatives. In its most common form, that of the marriage by the younger brother of the widow of the elder, it prevails very widely over the north and east of the couijtry, and is generally mentioned by ethnologists Tjinder its Biblically-derived title of the Levirate. It is less common in the centre and south. In the Panjab and upper portion of the North-West Provinces it has a special name, and the second marriage is duly accounted as such in the returns, diminishing thereby, to a considerable extent, the proportion of the widows. ■ But amongst the polygynous tribes of Assam, according to MrV Grait, the census Superintendent, the practice extends far beyond its Panjab limits. For instance, the heir to a Miri estate inherits with it the whole body of his father's wives, which are numerous in proportion to his means, an exception being made; it is well to state, of the actual mother of the successor. The system reaches its climax, some people might think, amongst the Garo tribe, where the bridegroom, by his marriage, binds himself to the reversion of —shade of Thackeray ! — his mother- in-law ! " Le monde s'ecroule," cries the hero of a recent popular drama," la belle-m^re reste!" There are several other topics connected with the marriage systems of, India, which are not only interesting from the point of view of the ethnologist, but also probably bound up more or less intimately with the wider questions involved in the consideration of this subject, with which alone the census has to deal. The matri- monial bargain, for example, is found in every stage of its progress, from the buying of a"" wife, which prevails to a great extent amongst the masses throughout India, to the more advanced tranS-action, in which the husband is bought, a practice in vogue amongst the higher classes only, in some parts of the country, and elsewhere extends to those below who aspire to social distinction by its .means. It seems probable, we repeat, that if the field occupied respectively by these two customs were ascertained, it might be found that there is some connection between the mode of contract and the age at which that contract is completed, and thus some , additional light thrown upon the distribution of sex, a subject which, as we saw in the preceding chapter, is sorely in need of such illumination. Then, furthermore, it would be interesting, in like manner, to learn how far the custom extends of the father's marrying his son below and his daughter above, his own rank, a practice connected with the question of dower and bride-gift respectively. Examples of this sort only show how the study of custom ought to be invoked in aid of mere statistics, in treatmg of wide questions of social importance, such as the marriage system. For the purposes of the present work, however, we must dispense with that aid, owinff to the fact, on which so much stress was laid in the opening chapter, that India is not so much a single country as a continent, so that the collation of such a mass of detail as is available 'in sporadic form on the subject in hand would entail the nreiDaration of a volume by itself. There are few topics that come within the scope of the census that 'are less susceptible of being dealt within in Imperial totals than the fia-ures ffiven in Table Ylll.'and its subdivisions, since the features common to all nrovinces are prominent, indeed, but subject to such remarkable variety m different tPrritorial subdivisions, that even a province is far too large a tract to be taken as a iiriit for analysis. In the reports from the Superintendents that have reached the Pfl-nfral Office up to the present time, a chapter is devoted m each case to the ccnsi- dSation of the local Statistics, and in some caees the information is supplem.ented by li 4 256 General features. the valuable experience of the reporting officer of the country concerned, an experience acquired by work and inquiry amongst the people themselves, and not merely derived from standard works of reference. Amongst those that come under this category, the chapters written in the reports on Assam, Bengal, the Panjab, and Madras are especially recommended for perusal, and to these may be added the corresponding part of the Burma volume, which deals with a marriage system very different from that of the rest of the Indian Empire. Some of the reports on the census of 1881, too, are perhaps even more valuable, since their authors had to work in virgin soil, and were, in this respect, the pioneers of their successors. In the present review, all that is proposed is to deduce, as far as possible, from the similarities in different parts of the country, the general type w^ith which the marriage system of the country at large tends to conform, and to indicate, under the same qualification, from the divergency in various directions, the results that seem to ensue from the adoption of that type. To begin with general features only, the following table gives the relative figures of the population in each of the three conditions for the 16 main territorial divisions of India, To make more apparent the difference between the Bast and the West in this respect, the corresponding return for a few European countries are added. Table A. — Showing the Distribution by Civil Condition of 10,000 of each Sex. Males. Females. Country. Unmarried. Maniecl. Widowed. Unmarried. Married. Widowed. India, 1891 4,873 4,647 480 3,389 4,851 1,760 „ without Burma - /,S52 /„66S 4S'0 S,33p J„8S3 'i,n§ 1. Madras ... 5,387 4,269 344 3,724 4,361 1,915 2. Bombay 4,514 5,021 465 3,027 5,273 1,700 3. Bengal 4,756 4,826 418 . 3,111 4,826 2,063 4. N.-W. Provinces 4,483 d,854 663 3,037 5,259 1,704 5. Oudh 4,557 ■1,891 562 3,160 5,239 1,601 6. Panjab 5,265 4,106 629 3.750 4,877 1,373 7. Sindh - 5,836 3,653 611 4,185 4,426 1,389 8. Central Provinces 4,668 4,890 442 3,560 4,960 1,480 9. Assam 5,622 3,968 410 4,138 4,162 1,700 10. Berar 3,838 5,588 574 2,635 5,769 1,596 11. Ajmer 4,839 4,653 508 3,355 5,094 1,551 12. Haidrabad 4,383 5,204 413 2,932 5,270 1,798 13. Baroda 4,323 5,158 519 2,997 5,517 1,486 14. Mysore 5,390 4,138 472 3,636 4,249 2,U5 15. Upper Burma 5,603 3,884 513 4.947 3,574 1,479 16. Lower „ 5,563 .3,980 457 5,136 3,927 937 1. England and Wales (1881) 6,193 3,463 344 5,928 3,314 758 2. Scotland 6,628 3,044 328 6,285 2,896 819 3. Ireland 6,871 2,750 379 6,344 2,698 958 4. Austria 6,151 3,554 295 5,771 3.416 813 5. Hungary 5,603 4,090 307 4,941 4,050 1,009 6. Sweden 6,233 3,411 356 5,958 3,233 809 7. Holland ; 6,237 3,392 371 5,940 3,317 743 8. Italy ; 5,989 3,610 401 5,397 3,672 931 9. Portugal 6,232 3,362 406 6,040 3,119 841 Now, one of the first points that will strike one in the above figures is that in Europe, omitting Hungary, which has a tinge oE Orientalism about it, at least 60 per cent, of the males, and of the females nearer that proportion than 50, are unmarried. In India, if we include Burma, which in this matter is less different than in others from the rest of the country as to its male population, no more than 48f of that sex are single, whilst the females in this condition come only just above the third. Burma so far as its single population is concerned, is remarkably similar in its ratio to Hungary, and shows a very much higher per-centage of unmarried women and girls than any other part of India. We may pass on to the married, leaving territorial diversity for future consideration. In Europe, on tbe whole, about one-third of the men, and -a slightly lower proportion of the women are paired off", whilst in India 257 the ■*?|>®^- ^®-^^- °* *^6 one, and about 48^ of the Other are in that ' state. As to .^^ widowed, the distinction between Europe and India, though marked' enough amongst men, where the proportions are 3-5 against 4-8 per cent., is extraordinary when we compare the figures for the other sex, since where there are eight widows in the temperate zone, there are not far from 18 in the tropics. Looking at the difference in the three conditions proportionately for the two continents, and using round numbers only, it may be said that India is behind Europe in its proportion of unmarried men by about 24 per cent. ; that it exceeds the latter in that of the married of the same sex by 46 per cent., and has relatively about 36 per cent, more widowers. As regards females, Europe is 43 per cent, ahead in the matter of spinsters, 4S per cent. 'behind as _ to wives, and at least 122 per cent, to the worse — numerically-— for widovvs. That is, men in India are largely, and women very largely, given to matrimony as compared with Europeans, with the consequence that in both sexes the widowed are in far higher proportion. The marginal table (B.) shows the distinctions in another light, namely, the average number of females to a Table B. — Showing the Sex Proportion in thousand males in the same civil each Condition. condition. In Europe, for in- stance, the unmarried women are usually less in number than the bachelors, though, except in Hungary and Italy, the deficiency is but trifling. In India, on the other hand, they reach but two- thirds of their possible mates. In the case of the wedded, the general excess of wives over husbands in Europe, not being explicable by polygyny, is attri- butable to migration of husbands, leaving the wives at home. It is also not an unusual occurrence for a woman to describe herself as married at the census, when, in fact, the bond is less durable. In India, as has been explained above, there is a certain amount of polygyny, and also of the concealment of wives. In a few cases, too, migration is un- doubtedly to be taken into con- sideration in connection with this point, as in Madras and Oudh, where the excess is con- siderable, and, on the other hand, in Assam, Ajm^r, and Lower Burma, in which the immigration is of married males only. The excess of husbands in Berar is attributed in part to under- enumeration, but the same feature in Haidrabad and Baroda has not yet been explained, and is certainly remarkable in its extent. The sex proportions amongst the widowed vary more than the rest, as might be expected from the different rules that prevail in India regarding the second -marriage of women who have lost their husbands In Europe, there seems to be an average of 2,500 widows to 1,000 widowers ; in India the mean proportion is about 3,570, which is considerably above that found in Hungary, the tract most abnormally situated in this respect, in the West. France and Germany have been left out of the •table purposely, since the figures for 1880 were affected by the losses_ of male life 10 years previously. It is worth notice in passing, that of the Indian provinces, Madras, Mysore, and Haidrabad show very high proportions, which may be in some 'degree attributable to the accidental cause of famine, a calamity before which, as -has been remal-ked' already, the so-called stronger sex seem to fall sooner than the Country, &c. No. of Females to 1,000 Males. Ud married. Married. Widowed. Total. India, 1891 670 1,005 3,533 963 „ without Burma • 663 j,oo: 3,66S 963 ■ Madras 707 1,044 5,702 1,022 Bombay 637 998 3,480 951 Bengal G58 1,006 4,962 1,006 N.-W. Provinces 625 1,000 2,375 923 Oudh 658 1,010 2,756 949 Fanjab 608 1,014 1,863 854 Sindh 596 1,006 2,260 831 Central Provinces 761 1,01^ 3,340 998 Assam 693 989 3,903 942 Berar 647 973 2,619 942 Ajmer 611 965 2,688 881 Haidrabad 645 976 4,204 964 Baroda 643 977 2,654 928 Mysore 668 1,016 4,433 991 Upper Burma 957 998 3,126 1,084 Lower „ 823 880 1,830 892 England and Wales 1,009 1.014 2,297 1,055 . (1861). Scotland (1881) 1,020 1,023 2,688 1,076 Ireland (1881) 963 1,023 2,640 1,043 Austria (1880) 982 1,006 2,882 1,047 Hungary (1880) 844 1,007 3,347 1,017 Sweden (1880) 1,014 1,005 2,413 1,061 , Holland (1879) 975 1,001 2,048 1,023 Italy (1881) 897 1,012 2,314 995 Portugal (1878) 1,058 1,013 2,263 1.092 78388. Kk 258 more patient. We have, then, Bengal, where the high ratio is due to the local marriage system, and Assam, where the same cause prevails throughout the more populous tract of the province. We have now elicited two main facts regarding the marriage system in India ; first, the abnormal prevalence of marriage, and, again, the abnormal inequality of the sexes in the condition of widowhood. To the discussion of these, then, with their causes and consequences, the rest of this chapter can be morg especially devoted. As regards the general tendency towards matrimony, it has first to be seen what are the influences that operate with less force in India than in Europe in repression of the mating instincts of humanity. The first is obviously the cost of the process. The charge of recklessness in regard to the assumption of the responsibilities ' of family life is too often made without considering the relativity of the imputation, and the only recklessness involved is on the side of those who cast the stone. Looking at the difference in the conditions, it is hardly straining language to say that the line between prudent and improvident marriage is more often overstepped amongst the masses in parts of the United Kingdom than it is in India, relatively, of course, to the respective populations. In the latter country, the material necessaries of life, measured in labour, cost but little. The standard of life, as distinguished from bare subsistence, although it shows signs of rising, is still remarkably low, and, as we have seen in the second chapter of this review, the question of building site and accommo- dation scarcely enters into the question, and clothing is much in the same stage. There are possibly cases, but they are so rare that they only serve to illuminate the general uniformity, in which the labour of a healthy adult does not sufl&ce to provide food for a wife and child as well as for himself. Then, too, at this point, another factor enters into the calculation. Throughout the masses, and it is with them alone that we must deal in this matter, the wife is in every . way the helpmeet of her husband, and the sooner she enters on her functions the better for his material comfort, Jiither she actually works at the family occupation or one of her own, or else a share of the domestic duties being taken by her off the hands of the bread-winner, he is enabled to devote more attention or time, or both, to his special avocation. This relief is, in India, no slight coiisideration, even though the duties in question be restricted to cooking, fetching water, and cleaning up the pots and pans used in those processes. Every section of the community, so far as general experience goes, seems to have some other with which it cannot mingle and whose touch is pollution to food or drink. There may be somewhere a bottom to the ocean of caste, and probably the European and Chinaman lie very near it, but, as a rule, the distinction is passed down as far as the inquiry is likely to follow. There is no field, therefore, for the public cook-shop, except in towns, or where Musalmans abound. The time and ceremonial spent in preparing meals is everywhere considerable and amongst some castes proverbial, in witness whereof we may cite the Hindustani saying, " eight Brahmans, nine cooking-places." It is thus essential that a man who is in the field all the day should have someone to cook and bring his food ; so, too, with the artizan, who has his loom, furnace, or tan-pit to attend to. The modern savage, says Mr. Eisley, wants a sturdy young woman who can carry his baggage, cook his food, collect edible grubs, and make herself generally useful, and mutatis mutcmMs, with a bit of the higher animal thrown in, this very fairly represents the whole duty of woman throughout the lower grades of Indian society. It will be seen that in these circumstances, the chief motive for prudential restraint is removed almost beyond sight. We come, next, to another side of the case, which was briefly touched upon in the third chapter. The ease with which a man can provide for a wife, the help and comfort he derives from her presence in his home, and the natural wish to leave behind someone to take his place in the world and carry on his name and tradition, all these, it may be thought, are incentives enough to matrimony ; but in India at least amongst the classes most strongly impregnated with Animism, to the realised expectations of this life is added the inducement of improving by a fertile union one's chances in the next. In Madras, writes Mr. Stuart, the strong tendency to marry has little to do with religious sentiment. This may be true, so far as the orthodox Brahmanic view of the destination of the childless adult is in question, but in the popular belief, the spirit of such an individual, so far from being safely encaged, to his own discomfort, in " the Hell called Put," is found to be exceedingly vicious and importunate in his visitations to his relatives still on earth. Burying him face downwards fails to keep him below ground, and when he has once got up. strings of leaves, and even silver nails, are insufficient to distract his inconvenient attentions. 259 Then, again, there is the more tangible ground of objection that if the childless man was in possession of any property, the application to the case by the courts of law of rules of inheritance framed entirely on Brahmanic principles, or on cusiom stereotyped under the neo-Erahmauio influences, might turn out detrimental to the next of kin. Therefore, we may say with Pahurge, it is evidently neces- sary that a man should marry^— in 'India, that is. Now, let us take the case of the womien, which, froni the Census point of view, is more important, and, sociolbgically, more interesting. In considering this part of the subject, we must first clear the Way by assuming or rejecting the hypothesis that was put forward in the last chapter, namely, that the deficiency in the return of women is due in some degree to an actual shortness in numbers, not solely to omissions from the schedules. If fewer girls, relatively to the other sex, reach a marriageable age, the less, of course, is likely to be the proportion of those left unwedded. The greater the diflPerence in age between husband and wife, the higher, in the ordinary course of events, is likely to be the number of the widowsl The examination of the circumstances in the case of the male section of the community, which has just been made, indicate how high is the probability that a girl, at some time or other of her life, will find a partner, and whether she is at a premium or subject to discount is a matter of conventionality only. Now, whatever the laxity of the hold that Brahmanic views on social questions, so far as men are concerned, may have on the popular mind, there can be no doubt as to the influence of such tenets in the case of women. It is superfluous to go over again ground already traversed in Chapter V., but it will be borne in mind what was therein laid down as regards the pervading atmosphere of Brahmanic notions, and how they had grown into a standard, as it were, of respectability in certain social features of the community. In marriage custom above all is this perceptible, for the conven- tionality on which it is based has its root amongst the sex to which, as in other countries, it is of the most importance, so it is by his conduct as regards marriage that the social aspirant is wont to be judged. It is noteworthy that it is in respect to marriage and widowhood that we can appreciate the advance of Brahmanism amongst the forest tribes of the Central Belt. It is amongst the highly orthodox tribes of the western Jamna basin that the Panjab Musalman preserves all his former customs. Further south, the Telingana population, which received Brahmanism in a fresher form than that which it assunled on its diffusion amongst the Tamils, surpass the latter in their orthodox adherence to the more striking parts of the Brahmanic system, and contain castes which in these respects are said to be more Brahman than the Brahmans. The more southern Dravidians, on the other hand, as Mr. Stuart points out. seem to have rejected all of the above-mentioned system that did not chime in with the most honoured of their own customs, and though orthodox in proportion to their position in society, the general level is decidedly below what strict adherence to the Puran would require of them. Now, it is time to see in brief what are the demands of orthodoxy in respect fco marriage. It is evident that in the earliest days of which we have even faint record in Aryan history, the girl was left to be courted pretty much as amongst the forest tribes of the present day. She could be won by prowess in athletic games, in war, or by gambling, and could likewise be gambled away or carried off by force or fraud, or otherwise disposed of. But, from what motive it is hardly necessary to inquire, the Brahmahical framers of the Puranic codes conceiyed a dislike to the position of an unmarried girl, and when they disliked anything they damned it with more than A.thanasian comprehensiveness. Thus we read that " the mother, father, and elder brother of a girl, who let her reach puberty unmarried, go to Hell," the precise section of the locality is not mentioned, but it is probably not the same as that reserved for the childless man. The girl was in a parlous state already, from which marriage alone could rescue her, so the act of omission''doubtless made it advisable to separate the family, at least by sexes.* ' Then comes the question of widowhood, on which much has been written, but nothing Prohibitio n^ed be quoted here, as the injunctions on the subject are apparently of still more modern of widow- origin than those respecting the marriage of daughters, and leaving out of question "^^^^E^- the Brahmans themselves, such precepts have no currency amongst the people, save what they derive from their value as rungs of the social ladder. But it is none the less apparent that the forbearance from marrying a widow is the first sign, in most parts of the country south and east of the Panjab, of the intention of severing oneself from a f di-mer unregenerate condition, and becoming incorporated in a higher section of the Brahmanic community. In many parts of Tipper India, where the Levirate is in « This was written before I had seen the controversy, on the subject between Drs, Jolly and Bhandarkar, with their exposition of the chances of misihterpretation of the te^t in question.— J. A. B. Ek 2 260 vogup, the thin end of the wedge is the reitriction of the marriage to the widow of one's elder brother, but amongst the masses of the people of Central and Southern India, apart from the explicable avoidance of half -worn articles Avhen new are to be had and their price is available, a widow of under 20 years of age has a very fair chance of a second husband, and Wellerism is rampant only amongst the very exclusive, or those who wish to be thought so. But we are speaking of the present day, and unfortu- nately the rise in the standard of living, which has been mentioned more than once in the course of this review as having ensued upon the state of peace and firm government of the last 30 years or so, has wi'ought a change in the views of those whom it affects, as regards matrimony. It has been said, on the one hand, that jamongst the well-to-do classes, such as Brahmans and "Writers, the taste for high living has Jed to the postpone- ment of marriage amongst the men, because they cannot get enough whilst quite JQiing to support a wife and keep up their standard of luxary, but against this it may be shown, with some reason, that if there be any signs of such postponement, it may be equally attributed to the rise in the price of food, and, in towns, of house rent, both of which hit the recipients of fixed salaries or cash payments, harder than the rest of' the community. At all events, the class in question is a very small one, and is by no means backward in letting the public have an opportunity of sympathising if the world does not go quite smoothly with it, so that if the inability to comply with any con- ventionality that the caste dictates had caused inconvenience, the foundations of the empire would surely have been shaken ere now, in the usual way. But a far more serious development of the Brahmanic system is indicated by some of the Superin- tendents in their census reports, and that is the strong tendency in the present day of peace and plenty to manifest their prosperity in the way mentioned above, firstly, by prohibiting the marriage of widows, and then by insisting upon carrying out strictly the Brahmanic injunction above quoted, and save themselves from the place to which the law-maker consigns them, by getting all their girls married before tbey have reached womanhood. Many cases have occurred within recent years to show that any movement amongst the literate classes in the direction of the abrogation of these two precepts isbutmouthdeep, whilst their heart is with the observance of them to the utmost. Longum iter jper precepta ; breve et efficace per exempla, but when opportunities occur for carrying into practice amongst themselves or their families some of the reforms they have been so strenuously endeavouring to impose upon others, it is remarkable to note what an amount of filial piety and of deference to the feelings of those to whom their respect is due comes into play, to prevent them from becoming martyrs to their principles. On the other hand, amongst the castes below the Brahman and E^jput, who have no education, and make no pretence of being sensible of any defects in their social system, we find continual attempts to conquer society by proving their claim to recognition through the adoption of these very tokens of high rank which, Ipj the admission of many who observe them, are blots on the social arrangements of the community, which it is the duty of men of light and leading to suppress. Thus, whilst the mouth is proclaiming its enlightenment and progress, the trunk is waddling backwards as fast as the nature of the ground will permit. Age, in con- This brings us to another section of our subject, and one which is, perhaps, the nection with most important, namely, the relationship between age and the general system on the marriage -y^j^ipij marriage is regulated in India. This can be exhibited in two forms. First, to put it tersely, the distribution of each condition by age, and, secondly, that of each age by civil condition. In anticipation of the final compilation of these statistics, and as an aid in the review thereof by the census Superintendents, the information collected at the enumeration of 1881, together with that for a good many foreign countries regarding which statistics bearing on this point were available, was reduced to a uniform proportional basis in the central office, and circulated to the officers concerned. It will be found, accordingly, that in all the Provincial and State reports, use has been made of this resource in dealing with the return for the special tract under review. It is superfluous to reproduce here the figures aforesaid in their entirety, as they cover about 10 pages of closely printed matter, but extracts will be given where relevant to the point immediately under discussion. It must first be remarked with regard to these statistics, that whereas those for European countries begin at the 16th year, it is necessary to carry back those for India to a far earlier age, for reasons that, if they, cannot be surmised from what has already been said in this work, will be abundantly illustrated by the selected proportions given in the margin on next page. Here we have the number in 10,000 of the single, married, and widowed respectively of each sex, who are less than 15 26i'^ years old. Whilst in the European countries, ©xceptr in Hungary, less' than '60 per cent, of the single of either < sex fall within this age^ ' in India more than tJaree- * quarters of the males and ; over 94 per cent, of the females have not^ attained > it. Put otherwise, in India only 23 out of every 100 bachelors, and only six out of the, same number of un- married women are over 15 years old. Then, again, as to the married, we find 5 per cent, of the husbands, and- 13-|- per cent, of the wives are beloW the age ■ at which the tables for,. European countries, except those of Austria and Hungary, begin. The pro- portions of the widowed, too, are very different in the two series. The male ratio, again, in India,' it will be noted, is higher than that of the juvenile widows owing to the more frequent remarriage of the former in riper years. "We can ' look at the figures at this age from another standpoint, namely, the distribution of the people of this age, according to their civil condition, irrespective of the relation they bear to their elders in the same cdudition. The following table- Table C. Proportion qnder 15 Years old of 10,000 in Country, &e. each Condition. . Males. Females. V i nd % O) ^ 4) ^ "S 1 ^ bo m IS ^ - a ^ ■.■'IBT-JJJ'I India - - 7,681 506 107 ,9,432 1,360 111 „ tvithout Burma 7,7/2 51 § m 9,5/4 1,391 113 Burma (5,752 4 2. 7,617 15 4 Scotland 5,802 5,536 _ Ireland - - 5,299 — " — 5,327 — — Germany 5,821 — — — - 5,956 — — Austria 5,631 — — 5,783 2 — Hungary 0,665 -— — 7,470 13 2 Holland 5,718 ^ri,(' „ — 6,i823 — — Italy 5,461 — — 5,866 — — France 4,937 — — 5,301 — — Sweden - - 5,449 — — 5,256 — — Portugal 5,761 — -~* 5,270 2 — Table D «d — Number in sach Condition of 10,000 of each Sex under 15. Proportion of , . Females per 1,000 1 Male's in Country, &c. Males. Bemales. each Condition. 1 ni ■a S t3 % ^ ? S « , ■g. t M h -§ 61' 1 ^ a" g ■ ^ t/2 s ^ 02 S ^ 02 — i — ., -^ ^ ^ India 9,390 590 20 8,247 ^ 1,702 51 822 2,703- 2,369 „ without Burma 9,S13 606 20 8,193 i,752 52 §17 ^,702 2,36§i ' B»™-{E ■ 9,997 3 — r^ 9,989 - 10 1 1,027 — i. — I 9,995 4 1 9,982 17 1 960 — — Scotland ■>! 1 r . 974 968 998 993 978 989 1,009 1,006 - 963 • '' — Ireland ■^~ —^ Germany y 10,000 — 1 , __ Holland Sweden France Austria- - • Hungary Italy - , - Portugal f. -: ■• > 10,000 — — 9,998 ^9,985"" 9,999 2 " 14 1 1:53 1:55 — J 9,998 2 968 1:8 . . . — ■ :' , Tii.,„^ ■.^^ .'o i-n -i o-*- +1.^ shows 'that comparison wim uuuuuixdo i^i a^u.xwj^v^.,.^, x.,. „^... .^^^^^^,^^^^^^^,^^ .^^ „^v,. age expluded from their tables, India returns nearly 6 per cent, of the boys and ] 7 perceBt of the girls as married, wi^h a certain proportion of widows and widowers. As before Hungary and Austria show a few girls married and widowed, and Italy still fewer showing that Juliet and Saint Elizabeth have their successors, but the highest of the numbers returned by these countries is insignificant The concluding section of the table simply compares the sexes as distributed between the three conditions in India. It shows that to every 10 husbands under 15 there are 27 wives, and to every 10 widowers aboiit 24 widows, whilst tlje unmarried girls would be Ifess by abqut' ot e : in every' feve. ■ -'■- Kk 3 262 The above three series of comparatiTe figures serve to set forth more prominently tii&ii those already given earlier in the chapter one of the piost remarkable points of difference^ between, the East and "West, that is, the prevalence in the former of juvenile marriage, with its complement of early, widowhood. There is a second point which is well worth notice here, namely, the prevalence of marriage at all ages, of which mention has hbea made both here and in a previous chapter. The following table, accordingly, shows the distribution of 10,000 of the unmarried of each sex within and beyond the period of which we have just been treating. is Table E. India. Seotland. Germany. Austria. Italy. Hungary. Age Period. CU i CQ |2( . 1 1 1 1 i 3 J" i 1 1 1 ■i Pi a 0—9 5,715 8,059 — — — ■ — ^ — — — - — — 10—14 1,966 1,373 — — — ' — — — — — — — Total, 0—U T,6S1 9,m S,S02 3,536 5,S21 3,936 5,631 5,7S3 546i 5,S66 6,663 7,476 15—39 2^170 510 3,768 3,642 3,834 3,523 3,915 3,538 3,959 3,483 3,169 2,382 40—49 84 28 201 319 159 209 212 288 253 266 90 78 50 and over 65 30 229 503 186 312 242 391 327 385 76 65 The figures speak for themselves. The average proportion of the unmarried males, which falls in Europe within the first 15 years, is contained in India in the first 10. The next period, 15 to 40 years, comprises what is in India the prime of life, ooiTesponding, no doubt, to 20 to 45 in Europe.' Allowing for the difference of age in reaching maturity and beginning to decline from it, the discrepancy between the two series is not excessive in ibis ses^--but in the next periods, omitting Hungary, the proportiori of 15achelors in Europe is remarkably Tiighef than "that in India,"bwing^ partly, of course,' to abstinence from marriage, and partly to the superior " staying power " of the European, after passing the prime of life. But when we look at the female side of the table, the difference is enormous, and no readjustment of the age periods will suffice to obscure the preponderance of married women in India at the reproductive ages, and, after 40 years of age, the almost entire absence, relatively speaking, of women who have never entered that condition. It is worth while, perhaps, to illustrate this fact a little more fully, by making use of a table corre- spondinff to the second of those given above in connection with the population of less than 15 years old. The following statement, then, shows by series at four age periods, the relative number in each of the three conditions, with the average number of females to 1,000 males of the same age and condition. Table F., showing the Number in each Condition of 10,000 at certain Age Periods above 15, and the relative Number of Females therein. « A.— Between 15 and 25 Years old. Country,,&c. Males. Females. Number of Females to 1,000 Males. 6 _ li I 1- i 1 6 03; 13 bo a CD 1 t3 India „ without Burma Scotland Ireland Germany Austria Hungaryi Italy Holland 4,879 9,276 9,660 9,649 '9j541 8,331 9,462 9,536 4,948 1 716 334 348 456 1,653 532 460 173 9 6 3' 3 * 16 4, 6 806 66s 8,623 9,107 '8,754 ' 8,^02 5,416 7,870 8,943 ,8,709 8M9 't 1,360 : 877 1,231 1,775 4,461 2,100 1,047 485 17 16 15 23 123 30 10 167 139 954 989 929 898 696i: 849 937 1,775 i,m 1,949 2,152 3,624 4,064 2,888 ■ 4,032 2,271 2,841 2,8i,8 5,030 7,298 ^^341 4,772 3,025 263- . ' ' ' ' .-' ' ■'-- B . — ^Betwee n 25 and 40 Years Id;' Numli 1 Country, &c. Males. Females. ec of Females to jOOO Males. .1 ■ 1 OJ T3 ■ : if >'' •'■i-- "'f" a ^ -§■■ , '^ 1 ^ -. f ' , , ^ , CO ^ _; '^ w. S &. K ,., h.^-^'U ..^ . < India , , 1,106 8,407 487 149 8,191 1,660 129 928 3,248 „ without Bwrma 1,091 §,/f2U 4S5 13>t S'.'/^p ^,677 111 929 3,302 Scotland 3,431 6,366 203 3^274 6,339 387 1,074 1,120 2,147 Ireland . 5,221 4,636 143 3,767 5,832 401 819 1,429 3,191 Germany ' 3,148 6,735 117 2,411 7,241 348 806 1,131 3,123 Austria 3,230 6,665 105 2,626 7,029 345 860 1,116 3,486 Hiingary . ■- 1,337 8,'SOI 162 741 8,475 784 563 1,014 ^,906 Italy - Uni 3,399 6,441 160 2,152 7,457 391 646 1,181 2,485 Holland 3,534 6,329 137 2,978 6,795 227 869 1,108 1,711 Country, &c. C. — Between 40 and 50 Years old. Males, ^ -§ Females. S Number of Females to 1,000 Males. ■c India without Burma Seotland Ireland ' Germany Austria Hungary Italy Holland 410 40^ 1,457 2,095 944 1,194 470 1,366 1,432 8,596 994 185 §,593 997 . 96 8,016 527 2,001 7,471 434 1,869 8,712 , 844 1,134 8.516 290 1,484 9,091 439 375 S,169 465 1,279 8,184 384- 1,496 5,612 .5,560 6,680 6,760 7,632 7,321 7,467 7,398 7,693 4,203 1,319 1,371 1,234 1,195 2,158 1,323 811 228 274 1,626 977 1,287 1,338 779 941 1,070 593 5§8 986 990 838 925 802 909 962 ; 3,916 I 3,961 2,963 3,460 3,838 4,437 4,797 2,856 2,164 D.— Fifty Years old and over. Number of Females to Males. Females. ■ 1,000 Males. Country, &c. «' 13 "3) r 1 ■ « 1 1 t ^ a 1 % ^ 1 ^ ■-4 . ^- ^ India 306 7,497 2,m 86 2,452 7,462 299 350 3,635 „ without Burma 302 7,49^ 2,200 16 ^,3^7 7,537 210 34.0 3,662 Scotland 1,140 7,069 1,791 1,961 4,282 3,757 2,236 787 2,726 Ireland 1,388 6,848 1,764 1,514 4,408 4,078 1,1*74 683 2,490 Germany 765 7,400 1,835 1,092 4,896 4,012 1,634 . : 757 2,501 962 7,482 1,556 1,388 4,909 3,703 1,584 720 2,613 Hungary Italy Holland 353 7,806 1,841 269 4,750 4,981 767 611 2,716 1,096 7,144 1,760 1,165 4,938 3,897 1,056 686 2,198 1,083 7,096 1,821 1,270 5,333 3,397 1,295 830 2,061 i The figures in the first period, that of the early prime, bear much the same testimony as those of the preceding one, though in a more strikmg aegree. Half the men are married and nearly 2 per cent, have already lost their wives. The women show nearly 5 ver cent, widowed and 87 married. In Europe about 85 per cent, are still single, and hot more than 15 per cent, are married. Hungary, of course, is exceptional. To an Indian statist there is always a sense of comfort m reaching tms 264- stepping stone between tlie primitive figures of the East and tlie dead level of sweet reasonableness found in the more regular returns of Western Europe. As regards the subsequent periods, it will be seen how the tendency in the case of males is much the same in the "West, as in India., though the degree in which it is manifested is far lower. The married men rise in proportion up to 50 years, and then decline in much the same ratio in both hemispheres. The Avidowers rise gradually to the end of life in India, but in Europe, though the course is the same, the rise, after 50 years, is more marked and more sudden. There is, of course, more divergence in the case of the other sex. The wives begin to decline at 40, | and after 50 only a quarter of the women are still mated, and the rest are widows, the proportion never married being negligable. In Europe, on the other hand, the proportion of wives follows that of husbands more closely, and ainongst the elders sciarcely two-fifths are widows, though the ratio of the unmarried is very much higher than it is in India at any age above 10. The results of the inequality can be appreciated from the sex proportions in the last three columns of each section of the tablej We iieed take little account of the single in connection with the point in question. As to the married, owing to the unions being early amongst both! sexes in India, the ratio of wives to husbands, between 15 a,nd 25, is only about 18 to 10 J which is far ' below that prevalent in Europe. In the course of the next period the Indian curves for the two sexes, if the details were set forth by diagram, would be seen to cross between 35 and 40, as many young widowers have remarried, whilst widows are not in demand. In Europe, taking account of later marriages, emigration, and the general tendency from various causes of males to be in defect at these ages, the balance is slightly in favour of the wives, and the proportion of widows is hot far below that found in India. Single women, too, are, as in India, at their lowest. We next find a very rapid fall in the latter country between 40 and 50, in the proportion of wives, with a slight one in Europe ; and the proportions in the widowed, too, correspond in relative variation, both in direction and degree. Finally, the same distinctions are accentuated in later life, and the divergence in India between the numbers of husbands and wives, widows and widowers respectively, is far greater than in the West, and the unmarried women everywhere, save in Hungary, outnumber the bachelors. This is all that can be said in the present work about the distinction between the marriage system of Europe and that of India, taking the latter, as a whole. We have now to devote attention to the diversity of practice found in different parts of that country, and, as the subject is one which involves the analysis of a very large amount of detail, far beyond the scope of a general review like this, it .is proposed to direct the inquiry chiefly towards the two most important questions only, namely, the relative prevalence of juvenile, or, as it is sometimes termed, infant, marriage, and its complement, the prevalence of widowhood, amongst the two sexes respectively ; for, as we have been shown in the tables already given, there is little of the nature of equality between the two. Men, as a rule, marry as early, and remarry, as their circumstances permit. The destiny of the women is regulated by custom, and, more distantly, by ideal conventionality, so that greater variety is found in their circum- stances than in those of the other sex. The tendency, however, such as it is, lies in the direction, as has been observed above, of earlier marriage and less widow marriage, according to the greater or less attraction exercised on the multitude by Brahmanic infl-uences. Eeviewing the whole field of statistics that is opened to us by the census, it seems to be established that in respect to juvenile marriage, and to an extent but slightly less, to widow marriage, the diversity of practice that appears in not only each Province and large State, but in every large subdivision of such units, is due, in the main, to local, rather Ihan to caste or social, custom. We see, that is to say, the lowest grades of the community in some parts of the country habitually marrying at an age below that of the upper ranks in others, not far distant and vice versa. An exception must be made, of course, to this rule in the case of Brahmans, who, though they may fall in with the local views as to the age at marriage, invariably prohibit, so far as our information is available, the marriage of widows. It will be seen in the comments on each Province, &c. that follow, how far these general tendencies are borne out, but it must be understood beforehand that the age quoted is, as in all cases where this factor enters, only approximately accurate. A few remarks on the territorial diversity at the principal age periods will not be out of place in connection with this part of the subject. The Imperial tables begin with the return of those under five years old. Of these, only a few over 103,0iD0 boys and 258,000 girls are said to be married, and about 7,000 of the one and 14,000 of the other are returned as widowed. The proportion of the married is 265 ttp'^^r.fif ^'^^ '"^ ?^'"°'?'' ^^^ ^^^ Gujarath States of Bombay, and for this there is Jnr«Wlil?'°" f ^^^,^y ^=^« SuperinWdents in 1881, that in a large local agricul- iSf +W V' "^^^^^^ °°°*^^°^ marriages only once in 12 years or so; every S.^SL>Sl°r ' ^^f^}'\ ^t the lucky moment is united to some other, under Tr£«/!^ T!l ^^'* *^^ "^'^^ '^y*'^^ ^''^°^1'^ ^^S^ too ^ate for caste propriety * ^ :«w!l? -S^. ' « sexes marry very early, not only in this period but in the next ff li -R ^^^^'^^^^ ,^ ^^,^ 10 comes the critical time when caste custom is supreme, at the ±5rahmamcal rules be oper^ttive. Here Berar comes in an easy winner, with °T+1,^r'" ■^''*:T,°i'*^ females wedded, but in the matter of boy husbands Baroda and the trujarath States of Bombay are its equals, on account of the continuance ot. the influence noted above. Haidrabdd comes hi^h as regards both sexes, a characteristic, also, of the adjoining tracts in the Madras and Bombay Presidencies ; ana /bengal though above the average in both sexes, is more remarkable for the number of its young bridegrooms. The period included between 10 and 1 5 is, perhaps, the most important of all as regards the marriage of females. It is unfortunate, theretore, that it is also that which, as we have seen in the last chapter, probably includes, to adopt an Irish figure of speech, most of the omissions. Taking the , ■ ^?^''™l^^ *^"^ ^*^^<^' Berar and Baroda again head the list of males, followed by the JNorth-West Provinces and Oudh, which in this respect overtop the average for the first time. Berar and its neighbour, Haidrabad, crown the female ratio with nearly three-fourths of that sex married, Baroda falling back below Bengal, Bombay, and Upper India. The community under 15, as a whole, such as is given in the preceding tables for all India, presents features which are more useful in considering the case of young men than of girls, for it includes far less of the physically adult in the former case than m the latter. Baroda and Berar are far beyond the rest in their proportion of the married in this period, the -North- West and Bengal following. It is superfluous to treat of the rest of the age returns in such detail as the above, so we may take, as the period of maturity, the years between 15 and 40. In this, the average pro- portion of the married is about 67 per cent, of the males and 85 of the females. Berar again heads the roll with 82 and 89 respectively. Oudh exceeds it as to females, but is far below. in respect to males. The ratio of married men is high, again, in the Central Provinces, Baroda, Haidrabad, Bombay, Bengal, and the North- West Provinces, and of married women in much the same area, substituting Sindh and the Panjab for Bengal. It is in connection with this period that the calculation of the fertility of the population is usually made. With the limited data in India this can only be a matter of conjecture. In England and Wales it is probably accurately given as 286 births per 1,000 married women of the reproductive ages. In India the Provincial Superintendents who have touched on this point estimate the figure, variously, between 159 and SlTf. At all events the general level of the com- putation indicates that excessive fertility is not a material cause of the increase of the population, so much as the alternative, that is, the high ratio of married couples. It is superfluous, looking at the limitations we have imposed on the treatment of the subject, to review the higher ages, beyond remarking that Madras husbands are the longest in maintaining their hold on this condition, and, omitting Burma and the Central Provinces, where the forest tribes, whose marriage system is very different, turn the scale, Sindh, Oudh, and the Panjab have the highest proportion of wives alive after 50. This is the more remarkable, as the husbands in these tracts present the lowest ratios in India. Against an average of 350 wives to 1,000 husbands at this age, we have 456 in Oudh, 439 in its neighbour, 393 in Sindh, and 363 in the Panjab. On the other side, in Haidrabad, only 268 are shown, in Madras, 287, Mysore, 290, and Bengal, 294. We paust revert awhile to compare with the above figures for the married, those Widow- regarding the widowed. Up to the tenth year the ratio follows, of course, closely ^°°d. that of the married, and Bengal, Berar, Baroda, and Bombay lead the way, and the North-West Provinces and Oudh occupy a higher place than they do with respect to juvenile marriage. The same items appear at the head of the next period, 10 to 15, when, in addition to the immature widows,„there is probably a considerable sprinkling of those who have been actually married in the European sense of the term. The highest ratio is found in Bengal, where it reaches 2| per cent, of the total number *This caste in Baroda returns as married 82| per cent, of the boys and 96| of the girls between 10 and 15, and ol| and 86f of those under 10. + jTaking the mean birth-rate at 47 ' 90 per 1,000, and the ratio of married women between the ages of 15 and 40 at 165 per 1,000, the average number of births to each married woman will be 290, or almost ' identically that of England and Wales, because, against the enormous birth-rate in India, must be' set off the 'remiarkable prevalence of marriage. T 78388. L ] 266 of girls of this age. In Berar it is 2'13, and Baroda and Uppei- India fall belo-w Haidrabad. At the riper ages, between 15 and 40, the widow is prevalent in Bfengal and Mysore to the extent of 15 per cent, of the women included, and in Madras and Assam to 12^ and 14 per centl respectively. Widowers, relatively speaking, abound mostly in Berar, and the North-West Provinces and Panjab. The number of Mdpws to widqwers is curiously varied. In Madras it is seven, and in Mysore nearly six times. In Bengal it reaches five, and in Haidrabad and Assam nearly four. Excluding territory much affected by male immigration, the North- West Provinces, Sindh, and the Panjab show the lowest ratio. It may be presumed that these jiropbi"- tions indicate roughly the inclination towards the marriage of widows in all but Madras, Haidrabad, and Mysore, where the losses of 1876-78 during the famine have not yet been effaced from the return. ^ -, It is now proposed to consider the returns territorially, using much the, same inaterial as has been set forth in the preceding remarks, but grouped anew for the main provinces, and illustrated by the comments of the local superintendent in cases Panjab. where his report has been received. We will begin with the Panjab, as a province which differs materially from the rest of India in some features of its. marriage system. In the first place, throughout the west of the province and amongst certain classes elsewhere, marriage is contracted on the same basis as in : Europe, that is, husband and wife begin to cohabit immediately after the ceremony, and there is no juvenile betrothal in those tracts, as is customary amongst a Brahmanic population. The latter prevails , in the east and centre of l^he province, and it is here that, juvenile marriages appear most frequently in the returns. There is here strong confirmation of the view set forth in a former part of this chapter that the prevalence of custom in marriage, except amongst Brahmans, depends upon considerations, not of caste, but of local example, for in the east of the Panjab, where Brahmanic influence prevails, even the Musalmans follow the practice of their Hindu neighbours, and m'ce /sersa in the west, as is shown by the marginal table. Here we have two districts, one per- vaded by orthodox Brahmanic views, the other under . Musalman influence. In : the latter the Musalmans scarcely marry their girls at all under 10 years of age, and the Hindus but rarely. On the other hand, along the Jamna, both religions show nearly 10 per cent.' of that age as married. The custom of Karewa, the Levirate, is found in mpst parts of the pi'ovince, except amongst a, few castes, such as Brdhmans and certain divisions of Eajputs; the Panjdb, accordingly, returns a smaller proportion of widows than most of the larger provinces of India, but still, as Mr. Maclagan points out, a quarter of the Hindu and a j&f th of the Musalman women of 15 years old and over are widowed. There is, n,o dqubt that Brahmanic disapproval of the marriage of this class has had some effect on the other religion, wherever the two are on more or less of niamerical equality. But, on the whole, the Panjab population marries at a maturer age, and is less strict as to widow marriage than most other parts of the country, though there is a great difference in this respect between the frontier and western plains and the more settled tracts of the centre and east. The latter tract resembles the general distribution of the popu- lation shown in the returns for the North- West Provinces, where the age at marriage is younger than usual. The examination of the returns for other provinces shows, however, the need of local experience in reviewing them, and as this is not at present available, there ig little to be said on the subject. There is a return of marriage by caste for this province, which shows remarkable differences in custom, but it has been furnished only for the province as a whole, and without knowing whether the peculiarities are local or not, it is unsafe to draw any conclusions from them, beyond a general remark that religion, as in the Panjdb, app^^ars to be but an insigniffcaht factor in regard to marriage custom where the caste is subdivided into a Musalman and a Brahmanic section, though the purely Musalman communities marry, as a rule, later than the bulk of the Hindu population. Bengal. We then come to Bengal, where Mr. O'Donnell has treated the subject in an interesting and valuable chapter, affording much food for reflection. As in the Panjab, we find locality the main factor in determining the age of marriage, but the tendency, to conform more and more to Brahmanic orthodoxy in respect to widow marriage is* N.W. Pro- vinces and Oudh. -. •. .. , KTo. of Married Girls under 10, to 10>000 Girls of ihat Age. Karnal. Shahpiir. Hindu Musalman 919 824 147 38 i 267 Subr Province. Number of married in 10,000 Boys under 10. Hindu. Mu^almdo, Animistic. Bengal' » Bihar Oriss^ Ohi^tia, Kagpur 46 i 67 ' — 926 323 — 42 — ' 36 287 1 - ;• 52 Sub-Province. Number of married in 10,000 Girls ijnder 16. Hindu. Musalman. Animistic. r N. Bengal <^ E. Bengal [ W. Bengal r N.Bihar - I S. Bihar Orissa . Chutia Nagpur 678 692 1,154 1,773 1,084 193 829 764 486 78^- 866 677 62' 118 very marked in this province amongst the castes that are on the borderland of respectability. The enormous population dealt with is subdivided by Mr. O'Donnell into large geographical sectious, by which means he shows more clearly how certain customs are purely local, not inherent in the castes that follow them. For exapiple, in respect to the juvenile marriages amongst boys the marginal table gives the propor- tions in the four great divisions of the province, from which we . _., . see how prevalent is this practice m Bihar, as compared to Orissa and Bengal, and how the Brahmanisation of the forest tribes of the Central Belt is making rapid progress. If we take the corresponding return for females, the territorial distinctions are still more marked, and, except in "Western Bengal and the two sections of Bihar, the custom of marrying girls below 10 years old is remarkably rare — for India. The contrast . between the Brahmanic and the Animistic sections of what is practically a uniform community in Chutia Ndgpur is very suggestive, as showing the insidious advance of the former in orthodoxy, as a means of establishing a better social position. But apart from the practical utility of demonstrating how localised is the custom in question, Mr. O'Donnell gives information which must, he considers, overthrow the time-honoured notion of the universality of the prohibition of widow marriage am6ngst the reputedly orthodox Brahmanic population of the lower Provinces. He finds the proportion of widows to the total female population of between 15 and 40 to be 26 J per cent, amongsb Brd-hmaos, where it is at its maximum so far as the prohibition iii question is concerned, so he takes this figure as the standard whereby to measure other sections of the community. By this criterion, it aippears that widow marriage is universal amongst the forest tribes of Orissa and Chutia Nagpur, where only from 5|^ to 7 per cent, of the women of the reproductive ages are shown as widowed. The Eastern Bengal Musalnians and the middle class Hindus of Orissa show but a little higher ratio, btit amongst the Musalmdns of Western Bengal, where the Hindus show the high &guxe of 24 per cent, tod over, the proportion creeps up. These deductions are certainly worthy of consideration, though it is, perhaps, inadvisable to adopt them without further scrutiny of the details. Tor example, it is quite possible that the paucity of widows at the age in question in Bihar may be as much due to the early agfe at which the boys marry as to the remarriage of young women who have lost their first husbands, since the proportion for Musalmans is higher than that for Hindus, and we have seen above that the former marry later than the latter in the case of both boys and girls, and the difference is greater amongst the boys, thus indicating the earlier loss of the husband. The detaals by caste which are given in this chapter of the Bengal report are very full and interesting. Mr. O'Donnell finds reason to attribute the earlier marriage of girls of the lower, or outdoor, castes, in some parts of the Province, to the feeling thait their life needs protection earlier than that of the upper classes, whose girls are secluded from harm in the zananah. The difference, however, between the figures for the two classes is scarcely wide enough for the assumption. The distri- bution of the proportions in various parts of the Province is unequal, and the fact that in most of the more prominent instances the protection of early marriage is extended to the hoys as well as the girls is against the hypothesis, whilst the relative paucity of widows is possibly due rather to the existence of the first husband than the acquisition of a sJecond. Mr. Stuart, like Mr. O'Donnell, has made great use of the tabulation of civil Madras, cdhdition by caste in reviewing this subject for the Madras Presidency, and from his ainalysis it appears that the marriage system in the south of India is subject to variations even more locally restricted than those of Bengal and' the Panjab. Taking the Province as a whole, it is chiefly remarkable for its high^proportion of the LI 2 268 unmarried of both sexes, for the low ratio of widowers, with a somewhat abaormally high one of widows. The unmarried are nob most numerous, relative to the total population, in the tracts where children bear the highest ratio, as might be expected, and the distribution seems to strictly follow local custom. The marginal table shows two extreme instances of this. In Tan j ore, with a low ratio of girls there is a high one of unmarried women, whilst in Yizagapatam, and, indeed, throughout the north-east coast tract, the reverse is the case. It might be assumed that to some extent Bengal custom had filtered down south, but the returns that Pei'-centage of Females. Under 10. Unmarried. Tanjore Vizagapatam 26-56 28-73 36 -70 29-26 were mentioned just above indicate that the proportion of unmarried in the Orissa castes, except Brahmans and writers, is remarkably high. The distinction seems to be, to a certain extent, racial, as in Bihar, as well as territorial. Eoughly speaking, the Telingana population is more married, if the phrase be permissible, than the Tamil or Malayalam, and the Uriya than the Telingana. In Malabar, as has been already mentioned, not only does polyandry prevail, but divorce is more in fashion, and the tie is generally more lax than in other parts of the Province. The diversity between the north and the south is due, presumably, to the greater influence of Brahmanic ideas amongst theTelanga population than amongst the more conservatively Dravidian Tamil, partly owing to the intercourse between the former and Northern India having been closer and of longer duration than that of the southern peninsula. The statistics regarding the female section of the community are those in which the diversity of practice is most apparent, even Brahmans varying between a per-centage of 5 of their girls married under 10 in the Tamil country and one of 12 in the South Deccan, and the figures for girl widows oscillating in harmony, between 10 and 35 in 10,000 of the age in question. Then again, the trading caste of the K6mti, who affect great orthodoxy in Telingana, show a ratio of 940 married and 26 widowed in 10,000 girls under 10, whilst the Chetti, the corresponding caste further south, is satisfied with 213 and 8. A still greater difference appears between, say, the Ambatban, or barbei'S of the Tamilian races, where the corresponding ratios are respectively 183 and 4, and the Mangala, who officiates in the same capacity amongst the more northern people, and follows the local tendency, by marrying off 762 of his daughters in the 10,000 before they are 10 years old, and finding 24 of them widowed by that age. By analysing and regrouping the caste statistics, Mr. Stuart arrives at the conclusion that Juvenile marriage, or the marriage of girls at an immature age, is "prevalent amongst 25 per cent, of the population, and fairly common lamongst " 15 per cent, more, whilst among the remaining 60 per cent, it is rare for a girl to " be married till she has attained puberty." By the same process of regrouping, the prevalence of widow marriage is summarised as follows : — " Among about 40 per cent. " of the population there is no restriction on the marriage of widows ; among about " 30 per cent, the practice is forbidden or rarely followed, while regarding the " remaining 30 per cent, the statistics are not conclusive one way or the other. We " shall not be far wrong if we assume that the marriage of widows is permitted and " practised by about 60 per cent, of the total population." To this may be added the remark that as in Telingana boys marry early as well as girls, the high proportion of widows there indicates how strong is the adherence to Brahmanic precept in respect to remarriage. On the other hand, in the breasts of the men of the Madras Presidency, hope triumphs over experience to a greater extent than in any other Province, for the proportion of widowers is the smallest recorded ; but they avoid mates in the same bereaved condition, as to every 1,000 widowers there are 5,702 widows. The losses during the famine must be credited with a good deal of this irregularity between the figures for the two sexes, but more is due to the advance of caste feeling fii Brahmanism. Mr. Stuart writes, " In nearly every considerable caste there are " some subdivisions which enforce the marriage of girls before puberty, and prohibit " the marriage of widows, and these subdivisions invariably hold a somewhat higher " position in public estimation on this account. It is easy to see that this encourage- " ment must produce a steady tendency towards the adoption of practices which are " in many respects injurious to the well-being of the people. And there is no " movement in the contrary direction. Once a caste has embraced the marriage '' customi- of the Brahmans it ncvtfr ivvcrts to its former position." 269 It is wortli noting, too, that the Borabay Karnatak, which adjoins the Telingana Bombay, tract of Madras, is also conspicuous for the early" marriage' of its females, and, to a less marked extent, for that of its males also. But G^ujarath is the division of this Presidency where boy marriages generally prevail, and here, again, from the caste returns prepared by Mr. Drew, it is clear that the custom is local, not governed by social position. In at least one caste the rules as to dower are so stringent that till within a generation or two ago girls were freely sacrificed, either at birth or during infancy, and even now there are grounds for suspecting deliberate neglect, if not worse treatment. In another caste the curious custom mentioned above prevails, according to -vyhich an auspicious season for marriage is only allowed by the tutelary goddess once in 10 or 12 years. Such rules are doubtless evaded, as by marriage to a dagger or a bunch of flowers, which are then disposed of down a well, and the youthful widow remarried at a cheaper rate, or the ceremony is gone through with the Pipal tree, which can never claim its marital rights or interfere with subsequent arrangements. But, as a rule, the custom of the upper classes filters through to the lower, and we find the village menials of Gujarath, for instance, marrying earlier than the landed aristocracy of the Deccan. The figures for Baroda give a very good notion of the special features of the former, especially in the caste returns*. The Sindh Sindh. figures partake of the nature of those for the "West Panjab. The age at marriage is far later than elsewhere, and there are more widowers and fewer widows. The age return is probably more inaccurate than usual, even at an Indian census. Crossing India, we have the r&turn for Assam to discuss. Mr. G-ait has examined Assam, the figures with some minuteness, and gives an interesting review of the marriage system in his Province. It appears that Brahmanic views as to the marriage of widows are confined in the Assam valley to Brahmans and writers, but in the Siirma valley, which is practically a corner of North-Eastern Bengal, only the lowest castes disregard them, and it is said that even amongst these, widow marriage is annually increasing in disrepute. The same remark applies to the Brahmaputra tract, though the population there is slower to move. The thin end of the wedge is found, not in the restriction of the permission to the widow of a deceased elder brother, as in Bihar, but in allowing only a virgin widow to remarry. As to the Hill tribes, we find very primitive marriage custom^ in force. The rule is generally to purchase, with the feint of a form of capture, but there are at least two tribes which do not purchase. The wives are invariably grown up, and so, of course, are the husbands. There are more women than men in most of the tribes, and men marry young, owing to the moderate price demanded for a girl. There appears to be little or no restriction, at present, on divorce or widow marriage, but as girls are plentiful, widows, except whilst yoUng, are not in demand. The position of/ woman in Burma is, as is well known, very much higher than Burma, that which she occupies in any Indian province. There is no juvenile marriage ; divorce is available to both parties alike, and is apparently more often initiated by the wife than the husband, though reunions of pairs once divorced are frequent enough to show that the process is not a very serious one. Life is easy, and a wife is more often than not a support, instead of being a burden. This being so, the law \a Burma, which probably follows at no great distance the popular sentiment, allows the wife a considerable control over the family property, and checks any tendencies towards polygyny, though the latter custom is not technically illegal. In old times, Mr. Eales points out, the Burman was a great slaveholder and polygamist. He seems to have been chastened by experience, but during the term of probation, if the law books are anything more than academic dissertations on what a monk thought ought to be done, man had left to him very little he could call his own. He could possess, if he liked, a head wife and a lesser, and also six concubines, but the law provided for the eight kinds of debts contracted by these without his knowledge, and, moreover, both wives and debts were subdivided into five orders respectively. Even supposing the head of the family had fortified himself by acquiring the knowledge of law requisite to circumvent his female connec- tions in these respects, it appears that he would not be safe from legal clutch unless he also studied what was prescribed in the case of a husband contracting debts without the knowledge of his two .wives and six concubines. In these circumstances, no wonder monogyny is the rule amongst the Buddhists of Burma. The g'eneral return for the lower division of the Province is disturbed by the large number of immigrants. Those both from the upper division and from India are generally married men without their wives, and amongst the women from India a good many are widows who have to earn their living by labour, along with their male relatives, who bring them across the LI 3 270 Mean age of the- married. Country or Province. Mean Age of the Married. . Difference Men. Women. Years. England and Wales 43-1 40 7 2-4 / Upper Burma [ Lower Burma 40 6 36 6 4-0 38 5 34 2 4-3 Madras 39 29 10-0 Bombay Panjab Assam 35 ' 31 38 2 8 8 27 25 28 7 8 7-6 6-0 10-8 Ber»r 351 26-5 8-6 India - 35-5 27-6 7-9 bay. In, Upper Burma the Buddhists show a high proportion of widows, which Mr. Bales attributes in part to losses during the operations against the dacoits, and partly to husbands having migrated and died in foreign parts. There is no prohibition of widow marriage, but apparently virgin brides are plentiful, so only the younger widows remarry. The proportion of the married of both sexes is low compared to that ruling in'india, not because there is less marriage in Burma amongst adults, but owing to there being no marriage at all before the young people are grown up. To start in life a young Burmese couple, as pointed out by Mr. Bales, wants but his da and cooking-pot. The universal bamboo provides the material for his house and furniture, his fuel and his water vessels, and when there is need, part of the family food. It is rare, accordingly, to find a youth of 23 years or thereabouts unmarried. It is noteworthy, also, that in Burma the difference between the mean age of Table Gr husband and wife is far less than in India, though it is still considerably above that found to prevail in England. The marginal statement must be taken with the qualification that the ages are inaccurate in detail, but as grouped for the purposes of this com- putation, not far from the mark on general lines. In order to show more clearly what is meant above by the influence of local or territorial fashion in regard to marriage, a table is given 'opposite in which the marriage statistics are worked up by religions, as well as by provinces. This point has been but cursorily mentioned in the foregoing review owing to the lax meaning attached to the term Hindu, when caste has not been tabulated in connection with this subject. In all cases, except in Burma and the Panjab, it is this form of religion that determines the provincial average, and it is curious to note in what degree the other religions vary from it. Taking the first section of the return, which groups the religions territorially, we find a few features strongly marked throughout. First, the Musalmans marry off their girls at a far earlier age than their boys and young men, and are by no means in such a hurry to get rid of either as the Hindus or Jains. Then, again, the line between the forest tribes and the orthodox is very distinct in the matter of juvenile marriage, and though there are a good many married girls of less than 15 years old compared to the boy husbands, both sexes show but, a small ratio in this respect, compared to that which prevails higher up in the social scale. The same class, too, with one exception, shows the lowest ratio of widows in the prime of life, whether the proportion of wives corresponds, as it usually does, or not. The Parsis are only given for Bombay,, as elsewhere they are both few in number and in many cases not thoroughly domesticated where they were enumerated. They show a very low proportion of juvenile marriages, and even in the prime of life abstain from matrimony to a greater extent than any other Indian community west of Burma. In the second part of the return the religion is shown in its territorial variation, and from the figures here given it can be judged how, with the general qualification as to the Musalmans and forest tribes just mentioned, the different communities tend to assinnilate to themselves the local colouring affected by the Brahmanic section of the population. The best example of this is found amongst the Animistic, who are more accessible than the rest, excepting, of course, the Jains, to the influence of Brahmanic precept, as they wish to rise in the social scale. Many of the Musalmans, it has been already seen, are but semi-converted Hindus, and change their habits the less as the antagonism between the two creeds is milder. The Christians follow the fashion as to juvenile marriage, especially in the matter of their girls. In Madras, where the men marry late in the tracts where Christian ^converts are most numerous, there are 704 brides of under 15 to 100 bridegrooms of the same age. In Bengal, where both sexes marry younger, there are only 254. In Bombay, the reverse is the case, and there are more girl wives in this community relative to boy husbands than even amongst the Hindus. It is probable that some of this discrepancy is due to temporary absence on the part of the youths in service, or on board coasting and 271 Table H. — Showing the Marriage System in different Religions, by Provinces. Under 15. Between 1 3 and 40. Province and Religion. Married per 1 ,000 oif each Sex. Wives per 1,000 Susbands. In 1,000 Males. In 1,000 Eemales. Wives per 1,000 lushands. Widows J ■ ; ! ' Males. Females^ Married. Widowed. Married. Widowed. ' per 1,000 Widowers. SECTION I. 1 Madras - - 15 97 6,369 611 18 806 125 1,628 7,216 Hindu , , Musalman ■" 'Christian Animistic 16 5 5 23 103 52 36 52 6,338 10,245 7,043 2,111 615 570 603 664 17 13 15 39 807 813 794 816 128 103 87 56 1,385 1,535 1,411 1,801 8,622 8,325 6,152 1,405 Bom'bay 67 235 3,306 757 31 871 101 1,085 3,075 . Hindu 'Musalman Jain Parsi Christian Animistic- - - 70, 42 56 10 11 17 248 141 227 32 44 65 3,310 3 125 3,572 . 2,809 3^50 3,7^1 772 682 644 507 434 763 31 32 35 16 16 23 873 872 852 696 832 897 103 92 139 52 43 42 1,079 1,171 952 1,281 1,001 1,208 t 3,131 2,623 2,817 2,976 3,104 1,898 Bengal 73 223 2,891 738 33 824 155 1,162 4,944 Hindu Musalman Christian Animistic 100 34 16 24 251 195 43 61 2,396 5,434 2,536 2,479 736 749 573 686 38 23 20 24 809 859 762 774 174 126 83 81 1,132 1,216 1,199 4,744 5,727 3,660 N.-W. Provinces 93 193 1,840 719 51 885 . 96 1,125 1,712 Hindu Musalman Jain 98 67 75 200 151 170 1,817 2,052 1,962 702 714 625 55 48 79 885 886 823 99 74 165 1,225 1,181 1,100 1,748 1,471 1,741 Oudh 83 165 1,802 724 38 905 71 1,131 1,764 Hindu - Musalman 88 47 171 126 1,750 2,466 724 730 38 35 906 904 73 55 1,172 1,781 Musalman Hindu Sikh 42 25 66 55 115 78 171 138 2,313 2,687 2,175 1,817 649 644 665 638 53 47 61 69 858 858 858 866 105 87 129 117 1,161 1,205 1,103 1,153 1,723 1,694 1,816 1,429 Burma, Upper Buddhist Animistic Musalman 1 1 1 1 3 5 — 528 529 543 561 35 36 31 33 587 585 712 651 89 89 70 102 1,191 730 480 2,725 1,283 1,292 Burma, Lower Buddhists Animistic Hindu Christian 1 9 2 1 3 13 2 — 559 520 544 592 495 38 38 47 22 26 680 673 692 850 650 61 61 60 52 49 1,023 1,141 203 1,108 1,364 1,140 334 1,575 Central Provinces Hindu Musalman Animistic 70 79 29 30 162 182 100 68 2,262 2,242 3,297 2,258 776 792 671 713 36 37 38 35 888 894 861 864 80 81 98 62 1,179 1,156 1,171 1,333 2,247 2,268 2,348 1,974 Assam Hindu - • - Musalman • Animistic 7 7 7 11 85 92 101 41 10,584 12,160 14,443 3,640 597 562 643 656 36 40 22 44 791 764 860 777 139 168 115 87 1,287 1,275 1,812 1,298 3,767 3,909 ,5,160 2,174 Haidral)ad Hindu Musalman Animistic 65 69 28 98 259 273 124 216 3,814 3,795 4,312 2,140 783 801 648 850 26 27 21 25 869 872 842 878 104 105 98 43 1,090 1,073 1,268 1,131 3,913 3,879 4,483 1,S42 Baroda Hindu Musalman Jain Animistic 135 143 74 83 23 253 269 144 ^ 93 j 45 1,732 1,735 1,799 1,074 1,935 749 757 683 634 661 36 36 36 36 25 888 891 869 843 878 81 ' 79 89 139 43 1,076 1,063 1,148 1,205 1,130 2,024 1,966 2,317 3,460 1,466 L 1 4 272 Table H.^-Showing the Marriage System in different Religions, by Proyinces---.77i lj406 1.391 1,040 898 701 655 Cfflitral Provinces 286 ago 194 a 10 310 34a 338 383 322 - 339 1,450 , i,S63 1,578 J.546 1,203 1,001 690 656 Assam 339 368 182 303 324 366 339 3S3 332 364; 1,516 1,684 1,606 J,S64 1,140 969 747 8S2 Haidrabad 268 298 220 as I 332 374 298 34a 311 328 1,429 I.S93, 1,S31 ,1.335 1,078 890 724 779 Mysore . . . - 239 ass 191 ao8 316 338 329 3SI 309 320' 1 1,384 M7a 1,364 1,414 921 835- 865 792 Central India 27B 3q6 142 161 236 253 2S0 340 293 333, 1,226 1,39" 1,449 1.453 1,188 939 798 730 Bajput&na - - - 328 3!lo 147 1S9 246 368 286 3a7 329 ■ 340 1,335 1,464 1,387 1.364 1,134 958 858 754 Baroda 314 344 163 186 273 316 288 340 289 334 1,321 i.Sio 1,424 1,407 1,108 933 866 77a Berar 310 333 177 aoi 283 3a8 276 334 264 392 1,310 1,48s 1,319 1,434 1,105 962 626 695 f Upper Burma 336 33s ,208 199 ?52 340 310 agS 232 ,319 1,327 5,a8S 1,229 1,168 1,199 i.ioS 887 954 (lower „ . . . 265 393 212 a4S 280 319 314 349 265 303 1,326 i,So8 1,259 1.38 1 1,190 1,179 927 1,070 India 326 347 173 188 287 319 318 .354 306 ,31? 1,409 1.52? 1,428 1,396 1,139 946 809 782 20-24. 25-29. 30-34. 35-39- 40-'44. 45- 49. 50-S4. ,55-59- 60 and over. Province or State. M. V. M. P. M. P. M. P. M. P. M. P. M. P. M. P. M. P. Madras 820 973 821 86$ 328 885 691 505 670 6611 365 305 427 460 177 IS7 616 6.3 'Bombay 844 939 968 931 862 851 641 573 629 626 383 342 416 435 180 16S I 420 538 .Sindh 785 875 900 928 968 964 562 461 612 640' 306 376 439 450 132 III 507 58a Bengal 702 838 341 895 807 818 648 568 624 608 369 333 394 410 169 170 469 610 (•N.-W. Provinces 878 918 868 88g 902 916 570 S50 713 724 833 317 470 507 144 ■45 467 609 (oudh 802 848 869 919 864 891 560 538 674 718 361 330 519 545 173 161 589 73' Panj4b 927 948 762 816 874 936 466 436 ' 596 675 333 395 458 532 132 143 629 S15 Central Provinces 707 833 840 911 923 904 626 498 756 674 2^0 236 464 446 95 100 4«9 633 Assam 757 928 882 928 865 83s ' 672 505 623 557 324 243 890 35 a 136 109 442 475 Haidrabad 807 983 970 916 914 932 538 441 738' 702 323 336 496 495 119 92 513 616 Mysore 860 943 901 904 329 825 700 6>3 656 601 439 378 418 44a 220 213 461 568 Central India 799 892 393 946 1,008 993 622 S81 764 738 321 277 438 4SS 115 112 .385 S02 Eajput^na 786 850 882 855 924 936 669 534 70S 749 338 308 472 504 171 158 441 566 Baroda 921 995 916 909 866 850 609 5S9 649 674 327 300 465 480 141 127 888 484 Berar 694 881 911 910 950 893 670 S45 768 668 330 288 539 480 166 123 678 634 /"Upper Buima (.Lower „ 869 844 789 766 728 708 684 626 568 567 I' 410 386 378 414 230 344 712 930 939 956 931 833 839 667 607 477 645 477 378 338 368 379 218 223 483 513 iHBIi 808 897 861 892 859 869 599 537 ,657 651 364 309 481 451 ,166 155 486 588 279 the oyder of sequence. Then, Numl)erof' Serial Order jf rovinees in. Age- in India, which the; Total. same Order is observed. 2 5 1 6 All. . 2 4 6 3 1 9 4 3 8 agadiiy the .return fo;r the other. sex is by no means uniform with that for males, as will be seen from the marginal statement, and it is the number of this sex, if there be any truth in the well-supported hypothesis started in a previous chapter of this worik, that is influenced more than that of males by physical circumstances preceding birth. The second, third, and fifth years alone of the girls agree in their sequence with those of the boys. There is no doubt that in some parts of India the year 1888-89 was a hard one for the masses, which might account for a paucity of births in 1889-90, but it seems scarcely probable that plenty was so abnormally dis- affect the ratio of the sexes in the births of the ensuing to attribute but a portion of the irregularity to fcributed in 1886-87 as to year. It is reasonable, therefore, actual ^variations in the numbep of births during the five years in question, and the Vest must, it is feared, be relegated to the unknown. Comparing the total falling within the above period, m the diflferent parts of the country, we find the highest j)roportion of children in the Panjdb, ^ind, and Assam. All three are very fertile tracts, on the whole, and as the two first named support a considerable burden of old people there is reason to suppose that the relatively high standard of living has something to do with the maintenance of the stream of population at its two extremes, In Assam life seems shorter, and the mean age is but 23*7 and 22'7 years in the case of the two sexes. The shortcoming of the female average is probably due partly to the number of the forest tribes, who work their women to death i at a comparatively early age, and in part to deficient returns, as there is a great' difference, far more than usual, in the results of the two enumerations in this province, and whilst the old men have proportionately increased, their mates have considerably; declined. On the whole, however, this result may be due to better enumeration at: more important periods of life. As regards the aged, Upper Burma almost reaches! European proportions, whilst the Lower division of the province is slightly below the average. This result is due in some degree, of course, to immigration, as it is in Upper Burma to the loss of men in the prime of life. Oudh and Berar both come high in the list in this respgct, probably on the same grounds as Sindh and the Panjab. We have next to consider the, period we have selected as indicating the prime of life, namely, between 15 and 45. In the case of Indian women it might be held advisable to place the higher limit nearly 10 years lower, but looking at the important part played by widows in the .domestic economy of the majority, it is not incorrect to stretch the period so as to include a large section of this class. The working period for men, too, begins lower than it would be correct to place its limit- in Europe, and ends earlier, but we are dealing with the prime of life only. Making allowance for these' qualifications, the marginal ' table, which gives the proportion of this period to 10,000 of each sex, shows the gene- rally high ratio through- out India. The adjust- ment of the tables in. the Panjab has undoubtedly depressed the figure for that province below its due in the ease of females, and to some extent in that of males, though there is no doubt a good deal of emigration on the part of the natives pf this province, and the ratio of the adjacent State of Kashmir shows that a high proportion is not to be expected. M m 4 15-44. Province, &c. 1.5-44. Province, &c. M. F. M. • I ■ F- Madras 4,558 4,672 Central India 4,400 4,595 J Bombay \ Sindh 4,749 4,710 Upper Burma .4,515 4,465 4,509 4,471 Lower Burma 4,788 4,480 Bengal 4,440 4,555 r N.-W. ProT. { Oudh 4,788 4,754 ' ' India 4,588 4,628 4,543 4,568 Panj?ib 4,315 4,466 England and "Wales 4,446 4,492 . Cent. Prov. 4,441 4,476 Scotland 4,413 4,451 A.ssam 4,546 4,605 Ireland 4,y07 4,408 Berar 4,614 4,592 France 4,498 4,464 Haidrabad 4,711 4,753 Germany 4,385 4,379 Baroda 4,826 4,759 Austria 4,488 4,525 Mysore 4,803 4,6178 .JIungary 4,564 4,680 Kashmer 4,400 4,595 Italy 4,457 4,566 Eajput^na 4,722 4,678 Holland 4,282 4,255 2^0 Computa- tions based on the age- returne. As regards the comparision witli the corresponding return for 1881, is must be borne in mind that we are dealing with very imperfect materials, so the deductions drawn can be but very general. The population, as a whole, is, no doubt, younger, owing to its replenishment by natural growth during the period of normal circum- stances since it was last enumerated. Excluding the Panjab, where the difference, as returned, is enormous, we find the greatest increase in Madras, which contains the worst of the famine-stricken tracts. Next comes Bombay, where the circumstances are similar. Oudh follows in much the same conditions, and Sindh is not far behind. Taking the whole of India, there are in every 10,000 males 91 more boys under five than there were in 1881, and in the same number of females, 107 more girls. It is curious that the ratio should have fallen in the C/entral Provinces, where there is no special cause for it, but as it was abnormally high in 1881, and the circum- stances of the population have since been prosperous, probably the difference must be sought for in the enumeration. In Bengal the decrease in the proportion is but slight. In Assam it is very large and must be classed with that in the Central Provinces. If we look at the other end of the table, we find the gradual recovery from famine marked in Madras and Bombay by the considerably in- creased proportion of the aged of both sexes. Sindh, Assam, and the Panjab, foi- some reason not immediately apparent, accompany an increased ratio of old men by a decreased one of the other sex. Bengal shows a small decrease, and both Oudh and the North- West Provinces an increase. On the whole, India is very much where it was 10 years ago in this respect. In the period between 10 and 15 our interest centres on the famine tracts. Mr. Stuart gives a useful table comparing these tracts with those better situated. An extract is given marginally for the period in question and needs no comment. Por Bombay, too, a similar table is avail- able, which is subjoined to that for Madras. The conclusions to be drawn from them are the same in both cases, and the selections are fairly parallel in point of prosperity and the severity of famine respectively. Even in so large an aggregate as the provincial total, the difference is very percep- tible. The great increase in the ratio of the young children in these tracts has been already described in the third chapter of this review. ^ There is little that need be said here regarding the age distribution of the different religious communities. In the provincial volumes for Madras, the Panjab, and Bengal, especially the last, the subject is discussed, but no conclusions of more than local application are derivable from the analysis of the figures. As a rule, the differences, where there are any, are due not to distinctions of religion but of locality, such as climate and food supply. In the case of Bengal, however, Mr. O'Donnell shows strong ground for holding the distinction between Musalmans and Hindus in Eastern Bengal in the matter of fecundity and length of life, to be due to social habits consequent upon change of creed. His analysis of the returns of age by caste is interesting and suggestive. In one point his remarks are certainly applicable beyond the limits of his charge, namely, as regards the fecundity and shortness of life of the forest tribes, though it is open to . question whether the latter is not partly due to the inability of such tribes to specify with accuracy any age over a certain number. It is not unusual for them to recognise nothing under a score, after about 35 or 40 years of age. The women of such tribes, on the other hand, undoubtedly die off very fast after all they go through in the course of their ordinary domestic duties and in childbirth. We must now pass from the statistics actually collected at the census to those which are merely based on them, supplemented by information derived from other sources, and by computation from the two combined. The object in view is to get ' the approximate value of life in different parts of India, so that it may be compared with that ascertained with greater accuracy in the case of the population of England and Wales, The first point to be investigated on these lines is the mortalityrate .in District. 1881. 1891. Madras. jyj r Grodavari (non-Famine) - ■ \ Cuddapah (Famine) 1,356 1,504 1,338 832 -.-, j Godavari ■ \ Cuddapah 1,077 1,366 1,075 728 Bombay. ^ r Khandesh (non-Famine) I Bijapur (Famine) 1,207 1,622 1,180 768 „ r Khandesh ■ \ Bijapur 1,035 1,409 1,028 651 281 Wl?rT J' ^^"^ '*'.'^'*^ *^^ °°™^1 ^a*« Of iBcrease, can be deduced the the diffAr^T^ro ^g^^^i/ssumng a stationary population, the expectation of life at would htrl ^''' ^^riP^^«^* *r"«* «ould be placed in the census data, all that Z W 1 7 ^"""i^ ^ *° '^"'P^^" *^^ ^"^b«r« r«*"r"ed at each age period at «.d ifv !5^T l""'*^ those -returned as ten years older' on the present occasion, dpd;n/+S'''''?''^^= .T ^^""-^ '^^" *° ^^ *^^ ™i^"*e influence of emigration, to eioedW ifr.! W *^^''^ ^'^l'"' t^ ^''^^'- Unfortunately, however, that simple expedient is not withm our reach. We have therefore to work on the mean popula- tion at each period as the base, and to go to other sources for the rate of mortality. ijucKily, there are two communities amongst whom the registration of births and deaths is conducted with special care. nmi««^n.*!'nJ'T'iw'li,^* ^^I^.^' ^f mentioned in the third chapter of this work, the Selected omissions ot both births and deaths is very prevalent. But in towns there is as is to registratior be expected, much greater accuracy than in rural circles. The difficulty, then, is to tind a town of which the component elements are not largely foreign and abnormal, which would provide a population sufficiently numerous to form, the groundwork of calculations applicable to any considerable portion of the population outside its limits. ±Jombay and Calcutta, for instance, are out of the question, since, as we have seen at page HZ, more than two-thirds of their inhabitants are recruited from outside. But Madras stands on altogether a different footing. Nearly three-fourths of its population was returned as home-born, and it is well known that this city, from its comparative want of manufacturing and shipping industry, does not attract the adult population of the surrounding country to anything like the same degree that its two sisters do. As the supervision of registration of vital statistics may be assumed to be equally efficient m all three towns, the results in Madras are obviously nearer those for a normal Indian community than those in either Bombay or Calcutta. We thus have m the birth and death returns for Madras the basis for computation of life- values applicable to the greater part of Southern India, and on them Messrs. G-. and H. Stuart have worked out the very interesting and instructive results that find a place in the Census Eeport for that Presidency. But lest the marked climatic and social distinc- tions between the south of India and the north, which have been insisted upon throughout this review, may have their effect on the age-values of the respective populations, it was advisable to seek for some data as nearly as possible corresponding to those for Madras in their approximate accuracy. These have been found in the returns collected in the North-West Provinces in connection with the supervision of the communities suspected of female infanticide. The registration of births and deaths among the youthful population of the tribes thus placed under inspection is remarkably good. The population is, no doubt, a special one, and from the very fact of being included in the scheme, must be, presumably, abnormal in its customs. But, on the other hand, the number of individuals in question is large and scattered, and the returns show that the practice of which they are suspected is only perceptible with difficulty and over, a very small area. Thus the data are, on the whole, fairly representative of a far larger population than that from which they are collected, and, in any case, they are clearly much more accurate than any procurable from other sources. It is also in their favour that they correspond on. the whole very fairly with the figures obtained from the Madras registers, ^nd are now available for a series of years that gives them a high value from a statistical point of view. The report on the Census of 1881 contained a valuable and suggestive chapter wherein the above data, so far as they were then available, were passed through the sieve of actuarial experience. On the present occasion the same course was adopted, for to whatever extent may be carried the analysis of an amateur in such matters, the results are certain to be inferior, in both matter and arrangement, to those obtained by the professional hand working on the same m.aterial. The information bearing on the subject was therefore made over, in its entirety, to Mr. Hardy, F.I.A., whose experience in 1881 could thus be extended over a wider field. Unfortunately for the completeness of this review, though probably Mr. Hardy looks at it from a different standpoint, he is no longer able to devote his full time to the reduction of the figureB and the calculation of the required values, so it has been found impossible to include his note in the present volume. It will be found, however, at the end of the second volume of the general returns.* * ]t must be mentioned, however, that when the computations were all but complete, and Mr. Hardy brought them to me in consultation, a considerable delay was caused by a suggestion I made regarding a re-adjustment of a portion of the figures for the Panjab, the early ages of which appeared to me to require such emendation. This put back, of course, the process of combination for the General Table, — J.A.B, r 78388, N n 282 Th e Life-tables dra-wn up by Mr. Hardy are ba^ed, like the mpst recent of those for England, on the results of the two last enumerations.. Tliat is, as the perio4 between 1881 and 1891 in India maybe considered to have been,, on the whole, a normal one, unaffected by. famine or excessive prevalence of epidemic .(Jisease, it is probably more accurate and useful to take it into consideration than, to ; attempt to find a " mean " which the comparatively frequent occurrence of disturbing factors is likely to render of no more than academic value. For the details of . construction and of the results, the reader must be referred to. the note itself, an 198,860,606 + i9.='94.5o9 111,082,159 101,324,656 + 9.757.5o3 107,072,956 63,459,819 54,932,908 + .8,526,911 32,805,690 28,574,662 + 4,231,028 30,654,139 97,535,950 26,358,246 + 9,537,006 + 4,295,883 221,172,952 66,006,763 198,860,606 54,932,908 , + 22,312,346 + 11,073,855 112,542,739 84,161,801 101,324,656 28,574,66a + ii,2i8,o83 + 5,587,139 108,630,213 31,844,962 97,S3S,95o ; . i ,v.(jW>»6i558ja46 ,, +■ 11,094,263 + 5,486,716 * The ai^rence of '43,716 between' ths population of i8c)i, shown here, and that given in Table I. is due to the exclusion of certain Bhfl tracts in Rajput&na, not fully enumerated on either occasion. a 2 IV III.— Towns and Villages classified by Population. Class and Population. INDIA. Pkovincks. States. No. Population. No. Population. No. Population. I. I— 199 343,052 32,625,858 248,564 24,279,022 94,488 8,346,836 z. 200 — 222,996 71,180,018 171,938 54,872,157 5i,o58 16,307,861 3. 5oo— 97,846 67,475,109 76,338 52,690,492 2i,5o8 14,784.617 4. 1, 000 — 38,128 51,349,338 30,260 40,741,875 7,868 10,607,463 S. 2,000 — 7,906 19,113,616 6,282 15,178,077 1,624 3,935,539 6. 3,000 — 3.770 14,059,089 2.977 11,100,332 793 2,958,757 7. 5,000— ijSoz 10,048,838 i,o5o 6,924,151 452 3,124,687 8. 10,000 — 366 4,402,062 Z20 2,659,579 146 1.742.483 9. 1 5,000 — i5o 2,541,135 94 i,6o2,5oo 56 938,635 10. 20,000 — 168 4,925,158 118 3,548,264 50 1,376,894 II. 5o,ooo — 76 9,309,434 60 7,550,172 16 1,759,262 fa. Travellers, Sfc. 56,334 26,331 3o,oo3 Unclassed-^ Lb. Not registered 1,589 137,442 — — 1,589 137,442 Total 717,649 287,223.431 637,901 221,172,952 179,648 66,050,979 Note. — The last item refers to the Bengal State of Hill Tipperah, where the villages were not classified. The number of towns of 5o,ooo, and over, is less by two here than in the next table, because the two suburbs of Calcutta are there shown in the population of the city though retaining their distinct numbering, whilst in this table they appear in Class 10. IV.— The Urban Population classified by Towns. Towns containmg a INDIA. Pkovinoes. States. Population of — No. Population. No. Population. No. Popiuition. I. 100,000 and over 3o 6,173,123 24 5,079,350 6 1,093,773 z. 5o,ooo — 48 38 2,589,686 10 665,489 3. 20,000— 149 4,492,113 III 3.417,347 37 1,074,766 4. 10,000 — 407 5,487,983 297 4,044,174 no 1,443,809 5. 5,000 — ... 896 6,164,900 577 4,023,339 3i9 2,141,561 6. 3,000 — 3oi 1,204,767 238 95o,6oS 63 254,162 7. Under 2,000 204 437,259 i3o 279,999 74 157,260 UnclaBsed (TVatie/fcr*, j-c.) - 35,856 — 6,629 — 29,227 Total Urban 2,035 27,261,176 1,416 20,301,129 619 6,860,047 v.— Variation in Population of Chief Towns. rj\r\-rxT-KT Population. Vakia- XION. m/^TX^'vr Population. Varia- X wv x-i . 1891. 1881. TOVVJ*. 1891. 1881. tion. I. Bombay* 821,764 773,196 + 48,568 34. Dacca 82,321 79,076 + 3,245 'Calcutta City* 681,560 433,219 + 284,341 35. Gaya 80,383 76,415 + 3,968 Suburbs 59,584 \ - . -a4,a75 69,642 251,439 29,982 5i,658 -191,855 + - 4,296 + 17,984 36. Amb^la* 37. Paizab&d* 38. Shahjehanpur* 79,294 78,921 78,522 67,463 71,405 77,404 + ii,83i N. Suburban Munict~ pality. S. Suburban Munici- pality. Howrah + 7,5i6 + 1,118 2. • 116,606 io5,2o6 + 11,400 39. Farakh4b4d* 78,032 79,761 - 1,729 Balli 16,700 7.037 + 9,663 40. R&mpur* 76,733 74,25o + 2,483 Total Calcutta and"! ( L Suburbs - J )78,370 878,541 + 99,829 41. Multan* 74,562 68,674 + 5,888 42. Mysore* 74,048 60,292 ' + i3,756 3. Madras* 452,518 405,848 + 46,670 43. Rawalpindi* 73,795 52,975 + 20,820 4. Hyderabad (Deccan)* - 415,039 354,962 + 60,077 44. Darbhanga 73,561 65,955 + 7,606 5. Lueknow* 273,028 261, 3o3 + 11,725 4S. Moradttad* 72,921 69,352 + 3,569 6. Benares* 219,467 214,758 + 4,709 46. Bhopal 70,338 55,402 + 14,936 7. Delhi* 192,579 173,393 + 19,186 47. Bhagalpur 69,106 68,238 + 868 8. Mandaiay*t 188,815 — — 48. Ajmer 68,843 48,735 + 20,108 9. Cawnpore* 188,712 i5i,444 + 37,268 49. Bharatpur 68,033 66,i63 + 1,870 10. Bangalore* 180,366 155,857 + 24,509 5o. Salem 67,710 59,63 1 + 8,079 II. Eangoon* 180,324 134,176 + 46,148 5 1. Jalandhar* 66,202 52,119 + 1 4,08 3 12. Lahore* 176,854 157,287 + 19.567 5z. Calicut 66,078 57.085 + 8,993 1 3. AUahibad" 175,246 160,118 + 15,128 53. Gorakhpur* 63,620 59,908 + 3,712 14. Agra* 168,662 i6o,2o3 + 8,459 54. Saharanpur 63,194 59,194 + 4,000 i5. Patna 165,192 170,654 — 5,462 55. Sholapur 61,915 59,890 + 2,025 16. Poena* 161,390 129,751 + 3i,639 56. Jodhpur 61,849 57,211 + 4,638 17. Jaipur 158,905 141.578 + 16,327 57. Aligarh (Koil) 61,485 62,443 - 958 18. Ahmedabad* 148,412 127,621 + 20,791 58. Mattra* 61,195 57.724 + 3,471 19. Amritsar* 136,766 151,896 - i5,i3o 59. Bellary 59,467 53,460 + 6,007 20. Bareli* 121,039 113,417 + 7,6" 60. Negapatam 59,221 53,855 + 5,366 21. Meerut* 119,890 99.565 + 19,8^5 61. Hyderabad (Sindh)* 58,048 48,153 + 9.89S 22. Srinagar*t 118,960 — — 62. Bhaunagar 57,653 47.79a + 9,861 23. Nagpur* 117,014 98,299 + 18,715 63. Chapra 56,352 51,670 + 5,682 24. Baroda* 116,420 io6,5i2 + 9,908 64. Monghyr 57,077 55,372 + 1.705 25. Siirat* 109,229 109,844 -' 6i5 65. Bikanir 56,252 43,283 + 12,969 26. Karachi* 105,199 73,560 + 31,639 66. Patidla 55,856 53,629 + 2,227 27. Gwalior* 104,083 88,066 + 16,017 67. Maulmain 68. Sialkot* 55,785 55,087 53,107 45,762 + 2,678 + 9,325 28. Indore* t 19. Trichinopoly* 3o. Madura 3i. Jabalpvir* 3*. Pesh&war* 33. MirzApur 92,329 90,609 87,428 84,481 84,191 84,130 83,091 84,449 76,847 75,705 79,982 85,362 + 9,238 + 6,160 + io,58i + 8,776 + 4,i09 — 1,232 69. Tanjore 70. Kambakonam 71. Jh4nsi*t 72. Hubli 73. Alwar 74. Ferozpur* 54,390 54,307 53,779 52,595 51,398 50,487 74,745 50,098 36,677 49,867 39,57c - 355 + 4.209 + 15,918 + 2,S3i ) + 10,867 _^ ... X : + Fieruree for 1881 no t available * Including Cantonments and Military Lines. VI VI.— Religion. Beligion. INDIA. PEoriNOBS. SlATBI. Hindu (^Brahmanic) . . - „ Arya „ Brahmo - - 207,68 8,724 39,952 , 3,o5i 155,129,941 39,014 2,988 52,558,783 938 Total Hindu Sikh - - Jain - - . . . Buddhist .... Zoroastrian - • - - Musahnan - - - * Christian - - - - Jew - - Animistic Minor Beligions - - - Beligion not returned 207,731,727 1,907,833 1,416,638 7,r3i,36i 89,904 57,321,164 2,284,380 17.194 9,280,467 i85 4^,578 155,171,943 1,407,968 495,001 7,095,398 76,952 49,550,491 • 1,491,662 14,669 5,848,427 i63 20,278 52,559,784 499,865 921,637 35,963 12,952 7,770,673 792,718 2,5*5 3,432,040 22 22,30O Total 287,223,431 221,172,952 66,050,479 VII.— Age. INPIA. PsOVINCES. States. Age. Males. Females. Males. Females. Males. Females. - 4,772,871 4,868,171 . 3,757,485 3,827,244 1,01.5,386 1,040,927 I 3,536,430 2,633,714 1,915,249 1,989,643 621,181 644,071 2 - - . - 4,203,315 4,468,349 3,239,481 3,455,796 968,834 1,012,553 3 4,650,886 4,968,439 3,612,038 3,858,675 1,038,848 1,109,764 4 4,467,878 4,471,446 3,409,284 3,424,241 1,058,594 1,047,205 0—4 20,631,380 21,410,119 15,933,537 16,555,599 4,697,843 4,854,520 5-9 - 20,908,467 19,578,667 16,121,870 i5,i3i,827 4,786,597 4,446,840 10—14 16,680,438 13,265,378 12,882,349 10,295,331 3,798,089 2.970.047 1 5 — 19 12,229,892 11,376,709 9,406,350 8,869,857 2,823,542 2,5o6,85» 20 — 24 11,743,796 12,577,424 8,961,462 9,669,199 2,782,334 2,908,225 25—29 12,828,365 12,680,985 9,742,243 ' 9,776,490 3,086,122 2,904,495 30-34 12,319,747 11,855,489 9,303,660 9,044,517 3,016,087 2,810,972 35—39 8,980,902 7,783,755 6,898,962 6,039,898 2,081,940 1,743,857 40—44 9,338,636 8,782,207 7,041,292 6,689,990 2,297,344 2,092,217 45—49 5,355,522 4,527,071 4,154,570 3,567,288 1,200,952 959,783 So— 54 6,015,467 5,970,618 4,551,100 4,574.183 1,464,367 1,396,435 55—59 2,618,702 2,378,652 2,061,145 1,900,719 557,557 , 477.933 60— 6,769,435 8,032,448 5,282,917 6,332,328 1,486,518 1,700,120 Age not returned 306,547 276,613 201,282 182,987 105,265 93,626 Total 146,727,296 140,496,135 112,542,739 .. 108,630,213 34,184,667 31,865,922 Vll Vlll.-Oivil Condition. Age and Condition. INDIAl *'•'*"-" -'S*'- a-'jj,. ■ :/• '-' Pkovinobs. Statbb. . . • J .a Mal«s. Females. Males. Females. Males. Females. 0-4- ■TTninarried Married Widowed 18,805,099 1 103,069 6,945 19,419,383 aS8,76o 13,878 15,851,537 ' 76,428 5,572 16,351,766 192,009 11,824 2,953,562 26,641 1,373 3,067,617 66,75 1 2,054 .Total 18,915,113 19,692,021 15,933,537 16,555,599 2,981,576 3,136,422 'Unmarried 18,344,180 15,663,196 15,50^,159 13,254,376 2,835,021 2,408,820 5-9- Married Widowed - 690,803 28,139 2,201,404 64,040 588,526 24,185 1,821,308 56,143 102,277 3,954 380,096 7,897 .Total ' - 19,063,122 17,928,640 16,121,870 15,131,827 2,941,252 2,796,813 'Unmarried 12,778,801 S,977,3oi 10,802,316 5,003,591 1,976,485 973,710 10-14- Married 2,342,433 6,016,759 2,017,614 t 5,137,870 324,819 878,889 Widowed 71,471 174.532 62,419 153,870 9,052 20,662 ■total 15,192,705 12,168,692 12,882,349 10,295,331 2,310,356 1,873,261 'Unmarried 13,622,958 2,133,452 ,,1' 11,455,088 ,1,738,832 2,167,870 394,620 iS-34- Married Widowed 29,683,040 1,350,699 38,144.713 4,160,^48 24,795,495 1,163,132 32,oi3,8|i3 3,607,418 4,887,545 187,567 6,i3o,90o 5'53,i3o .Total 44,656,697 44,438,713 37,413,715 37,360,063 7,242,982 7,078,650 'Unmarried - . - 1,025,383 i .209,539 864,551 1 5 8,647 160,832 50,892 35-49- Married Widowed 18,649,824 1,858,073 ia,o34,o5a 6,996,592 15,656,068 1,574,205 10,155,572 5,982,957 2,993,756 283,868 1,878,480 1,013,635 .Total - 21,533,280 19,240,183 18,094,824 16,297,176 3,438,456 2,943,007 'Unmarried 430,603 128,749 362,724 97.S03 67,879 31,246 Married 5o and over - Widowed 10,637,733 3,087,748 3,688,667 11,224,933 8,922,138 2,610,300 3,134,434 9,575,293 1,615,595 477,448 554,233 1,649,640 = , Total 14,056,084 15,042,349 11,895,162 12,807,230 2,160,922 2,235,119 ^Unmarried - 129,405 100,413 92,052 71.841 37,353 28,572 Age not re- turned. Married Widowed 113,398 9,408 io4,Sgi 22,906 81,860 6,703 75,023 1 5,200 31,533 2,705 29,568 7,706 • .Total 252,211 227,910 180,615 162,064 71,596 65,846 fUnmarried 65,136,429 43,632,o33 54,937,427 36,676,556 10,199,002 6.955,477 Total return- Married ing Civil -^ condition. Widowed ' 62,120,300 6,412,483 62,448,946 22,657,449 52,138,129 5,446,616 52,530,029 19,402,705 9,982,171 965,967 9.918.917 3,254,724 LTotal 153,669,212 128,738,408 112,522,072 108,609,290 21,147,140 20,129,118 Not enumerated by Civil Condition.* ""■' 13,058,084 11,757,727 20,667 20,923 13,037,417 11,736,804 Total Population W6,t27,296 140,496,135 112,^2,739 108,630,213 34,184,557 31,865,922 •A. Assam (N. Lushai) -' - 20,667 20,923 - — . IT 'Kaship^f.''""' • ■■ - — - — — 1,363460 W90.6S3 '■ 1-.^ BajputSnfe - -/ - -- — — 6,842,407 ; iM6.919 B.- Central' Inflii - - — — — 5,831,271 4.88o,SlS. Mahi KMittik (Bombay) - • - - -^ — — — 10,5!i9 B.6S1 Total States - - - - 13,037,417 11,736,804 INDIA - 13,058,084 11,757,727 - - - - a 4 vm IX.— Instruction. INDIA. Peovincbs. States. Total. Males. Females. Males. Females. „ ,Males. . ; Females. "CTnder Instructian 3,195,220 2,997,558 197,662 2,593,887 162,248 403,671 35,414 Literate 12,097,530 11,554,035 543,49s 9,903,664 447.9*4 1,650,371 95,571 < Illiterate 246,546,176 118,819,408 127,726,768 99,797,906 107.794.481 19,021,502 19,932,287 Total returniug In- ^ struction. 261,838,926 133,371,001 128,467,925 112,295,457 108,404,653 21,075,544 20,063,272 'Under Instruction 2,695,792 2,518,240 177,552 2,170,695 145,756- 347,545 31,796 a Literate 763,120 67.5,357 87,763 580,210 72,588 95,147 15,17s ft a Illiterate 99,468,421 49,960,163 49,So8,258 42,169,671 4i,74».733 7,790,492 7,759,525 .Total 102,927,333 53,153,760 49,773,573 44,920,576 41,967,077 8,233,184 7,806,496 'Under Instruction ■ 441,224 427,595 13,619 379,023 11,484 48,572 2,145 * T, Literate 3,052,787 2,864,980 187,807 2,460,633 157,797 404,347 3o,oio 1 < Illiterate 40,426,673 18,574,203 21,852,470 15,518,799 18,359,483 3,055,404 3,492,987 .Total 43,920,684 21,866,778 22,053,906 18,358,455 18,528,764 3,508,323 3,525,142 'Under Instruction 58,204 51,723 6,481 44,169 5,008 7,55* 1.473 o Literate 8,281,623 8,013,698 267,925 6,«62,821 217,539 1,150,877 5o,386 n Illiterate 106,651,082 50,285,042 56,366,040 42,109,436 47,686,265 8,175,606 8,679,77s .Total 114,990,909 58,350,463 56,640,446 49,016,426 47,908,812 9,334,037 8,731,634 Not enumerated by education.* 25,384,505 13,356,295 12,028,210 247,282 225.560 13,109,013 11,802,650 Total Population 287,223,431 146,727,296 140,496,135 112,542,739 108,630,213 34,184,557 31,865,922 '•Assam ^N. Luthai) ■ — — — 20,667 ao,9J3 _ _ BeagaiiChiMagmg HUU) - - - 69,566 4Mao - — *A.- Madras ( YizagapatamAgency ) — — — 1«7,0« 146,917 - - Total, Provinces - - - 347,882 235,560 - - 'Kashmir - - - - - - 1,363,180 i.igofiii fiajputtaa — — — — — 6,34SA>7 ifiiim * Centrallndia — — — — - 6,831,271 4,88o,JiS B.- Hill Tipperah (Bengal) - — — — — ■ - 71,6»6 6S.846 Mahi Kintha (Bombay) - — — — — — 10,679 8,657 Total, States - - - - - - 13,109,013 11,803,660 INDIA 25,384,505 13,356,295 12,028,210 — - - - IX X.— Parent Tongue. (a) Languages by Linguistic groups. A. Aryo-Indic B. Dravidian C. Eolarian D. Gipsy Dialects E. KMsi F. Tibeto Burman G. IKdn Annam H. Taic, or Shdn FAMILY - -I J. Maylayan K. Sinitic L. Japanese M. Aryo-Eranic - N. Semitic 0. Turanic P. Aryo-European Q. Basque R. Hamitic or Negro - Language unrecognisable Return left blank Total enumerated by Parent Tongue Population not enumerated by"! Parent Tongue " J Population returning. TOTAL (b) Languages, in order of pre- valence in the tracts enumerated. (1) 1,000,000 and over each. I. Hindi— A. i. Bengali — A. 3. Telugu— B. 4. Marathi — A. 5. Panjabi — A. 6. Tamil— B. 7. Gujarathi— A. i. Kanarese — B. 9. Driya — A. 10. Burmese — E. 11. Malayalam — B. iz. Urdu {Masalmdni)* — A. 1 3. Sindhi— A. 14. Santhal — C. 1 5. Western Pahari (Panjab)— A. 16. Assamese — A. 17., Gond — B. 18. Central Pah4ri (Garhwali, &c.)— A. 19. Marwadi — ^A. 20. Pashtu — ^M. - - " (2) 100,000 to 1,000,000. ai, Karto— K. 22. MiSuda (K&l)— C. 23. Tulu— B. 24. KachH— A. aS. Gypsy dialects (Lambtai, &c.)— D. 26. Oraon — B. - - - 27. Rhand (Khdnd)— B. 28. English— P. 29. M6n (Talaing)— G. 30. Bal6ch— M. 195,463,807 52,964,620 2,959,006 401,125 178,637 7,293,928 229,342 178,447 4,084 713,350 93 1,329,428 55,534 659 245,745 1 9,612 363 19,659 262,047,440 25,175,991 287,223,431 85,675,373 41,343,762 19,885,137 18,892,875 17,724,610 15,229,759 10,619,789 9.75i.8»5 9,010,957 5,926,864 5,428,250 3,669,390 2,592,341 1,709,680 1,523,098 1,435,820 1,379.580 i,iS3,384 1,147,480 1,080,931 674,846 654,307 491,728 439,697 401,125 368,222 320,071 238,499 226,495 219,475 Languages, in order of prevalence in the tracts enumerated— continued. 3i. Kacliari(Bi5do)— F. 32. Nipali Hill dialects— P. 33. K6rwa (Kur)— C. 34. Khasi— E. 35. Shan— H. 36. Bhil— C. 37. Garo — F. 38. Halbi— A. 39. Khvin — F. 40. Tipperah — F. 41. Naga dialects— F. 42. Sawara — C. (3) 10,000 to 100,000. 43. Mech— F. 44. Mikir — F. 45. Kathe (Manipuri) — F. 46. Kharria — C. 47. Arabic — N. 48. Baiga — C. 49. Lushai — F. 50. Laliiug — F. 5i. Chinese — K. 52. Goauese and Portuguese — A. 53. Kodagu (Coorgi) — B. 54. Abor (Miri; — F. 55. Mal-Pahadia— B. 56. Gadaba — C. 57. Kashmeri — A. 58. Brahui— B. 59. Persian — M. 60. Eastern Pahari (Nipali, &c.) — A. 6i. TibeUD— F. 62. Kuki— F. 63. Minor Kolarian dialects (6) — C. 64. Lepcha — F. (4) 1,000 to 10,000. 65. Bhutani— F. 66. Negro dialects — K. 67. Kanawari — F. 68. K6ch— F. 69. Minor Dravidian dialects (3) -B. 70. Kakhytn- F. 71. Minor Bodo dialects (4) — F. 72. Mahl— B. 73. Khamti— H. 74. Palaung~G. 75. Malay — J. 76. German — P. 77. French — P. ^ - 78. Hebrew (Israeli') — ^N. 79. K6ta and T6da— B. 80. Sal6n— J. 81. Minor North-Bast Frontier dialects (3) — F. (5) Under 1,000. Indian languages (5) Other Asiatic languages (8) European languages (23) Population returning. 198,705 195,866 i85,77S 178,637 174,871 148,596 145,425 143,720 126,9x5 121,864 102,908 1 02,0 39 90,796 90,236 88,911 67,772 53,35i 48,883 41,926 40,204 38,5o4 37,738 37,218 35,703 3o,838 29,789 29,276 28,990 28,189 24,262 20,544 18,828 11,965 10,125 9.470 9,612 9.265 8,107 7.65 1 5,669 4.314 3,167 2,945 2,847 3.437 2,2l5 2,171 2,171 ",937 1,628 1,282 5x3 i,i95 2,913 V 24230. *UrdA is returned as a separate dialect in Southern, Western, and Central India only. XT.— Birthplace. Enumerated in the Province o* State. BOEK IN THE PeOTINOB OB State. Province or State. TOTAL enumerated. Born in the Province or State. Born in . contiguous Territory. Born elsewhere. TOTAt. Enumerated in contiguous Territory. Enumerated elsewhere in India. Ajmer-Merwara 542,358 433,664 83,447 25,247 490,221 52,720 3,837 Assam 5,476,833 4,961,324 419,562 95,947 5,017,019 53,417 2,278 Bengal* 71,345,947 70,597,556 5o8,8o5 239,486 71,838,732 1,191,123 49,953 Berar 2,897,491 2,446,576 387,198 63,717 2,561,430 113,498 1,356 Bombay* 15,978,463 14,882,954 8o5,3o2 290,207 15,985,829 905,540 197,335 Sindh* 2,871,303 2,665,691 171,254 34,358 2,692,973 1 8,940 8,342 Burma* 7,602,429 7,273,817 150,649 177,963 7,282,348 2,416 6,iiS Central Provinces 10,784,294 10,289,405 420,239 74,65o 11,313,422 946,870 77,147 Coorg* . - - 173,010 120,110 51,628 1,272 122,772 i,4i3 249 Madras and Laccadives* 35,326,412 35,029,729 272,398 24,285 35,646,453 412,669 204,055 North- West Provinces* 34,249,210 33,100,077 1,118,045 3i,o88 34,768,181 1,459,124 208,980 Oudh* 12,649,049 i2,i79,i3o 444,645 25,274 12,705,441 389,665 136,646 Punjab* 20,866,787 19,780,050 1,008,424 78,3i3 20,608,421 729,504 98,867 Provinces 220,763,586 213,760,183 5,841,596 1,161,807 221,033,242 6,277,899 995,160 Hyderabad* 11,535,480 11,151,767 301,544 82,169 11,539,750 375,405 12,578 Bardda* 2,415,197 2,103,474 289,852 21,871 2,355,870 246,757 5,639 Mysore* 4,941,882 4,746,084 i83,652 12,146 4,889,617 136,694 6,839 Kashmer* 2,527,056 2,+57,799 66,346 2,911 2,548,726 87,545 3,382 Rajput&na 11,986,196 11,578,814 370,832 36,55o 12,478,431 745,607 154,010 Central India* 10,317,242 9,564,629 680,467 72,146 10,146,485 558,197 23,659 Bombay States* 8,058,299 7,347,901 609,1 1 5 101,283 7,950,357 480,184 122,272 Madras States* 3,700,369 3,596,784 98,910 4,675 3,629,824 32,338 702 Central Province States 2,160,511 1,554,211 572,069 34,231 1,579,582 25,264 107 Bengal States* 3,296,125 *.767,707 513,840 14,578 2,878,968 12,118 99,143 North - Western Province States. 792,491 724,327 65,149 3,01 5 788,135 59.43- 4,371 PanjAb States* 4,263,192 3^61,338 480,375 21,479 4,206,065 443,425 l,3o2 States 65,994,040 61,354,835 4,232,151 407,054 64,991,810 3,202,971 434,004 Total retumirrg birth- place. 286,767,626 275,115,018 10,073,747 1,568,861 286,025,052t 9,480,870 1,429,164 Aden, Quetta, Sfc. excluded - 89,921 — — — __ Not enumerated by birth- place. 375,884 — — — - — ~ INDIA 287,223,431 — — — — — ■— * Those not returning birthplace are here excluded. fThe total number of persons not bom in India, including the Ifr^nch and Portuguese possessions, was 661,637. Of these 478,656 returned as their birthplace countries contiguous to India; 60,519 countries in Asia remote from India iucludins however, China ; ioo,55i the United Kingdom ; 10,095 other European, American, and Australasian countries ; whilst 11 816 were born in Africa, &c., or at sea. ' XI XII.— Infirmities. ■iaqL- A.-INDIA. Age. Insaitb. Dejjt Mxns, Blinii. Lepees. Males. Females. Males. Females. Males. Females. Males. Females. o— 4 5-9 10 — 14 774 2,673 3,728 52Z 1,617 2,338 5,452 17*086 15,758 4,o57 10,969 8,787 9,426 14,836 ^ 14,842 6,368 9,521 9.437 426 843 2,285 »85 607 i,3oS iS — 19 20 — 24 s5 — 29 30-34 4,^93 4,791 5,599 5,741 2,758 2,884 2,823 3,145 12,968 11,660 10,814 9,914 7,845 7,263 6,569 6,123 -13,458 13,911 14,472 15,156 9,388 10,292 11,860 14,217 3,857 5,569 8,342 11,431 1,934 2,277 2,867 3,678 35—39 40—44 45—49 4,833 4,482 2,543 2,461 2,770 1,688 7,282 7,498 4,552 4,175 4,8o3 2,793 12,816 1.5,985 12,914 12,276 17,274 13,394 11,495 14,474 9,488 3,090 4,014 2,403 50—54 55—59 60 and over 2,557 1,264 2,673 2,o5o 903 2,558 5,487 2,9.53 8,873 3,698 1,909 7,25o 17,163 14,369 59,772 20,378 i6,oo3 78,999 11,065 4,687 11,131 3,068 1,415 4,o23 Age not returned 188 123 200 123 233 2o8 125 60 Total 45,639 28,650 120,497 76,364 229,253 229,615 95,218 31,026 Ins B.— PROVINCES. Age. ANE. Deaf Mute. Blito. Lepsbb. Males. Females. Males. Females. Males. Females. Males. Females. 0—4 5-9 :o — 14 i5— 19 20—24 15 — 29 30—34 35—39 40—44 45—49 50—54 55—59 60 and over Age not returned 633 2,103 3,005 3,520 8,983 4,699 4,845 3,714 3,790 2,183 2,139 1,094 2,288 107 412 1,292 1,916 2,273 2,398 2,427 2,689 2,149 2,378 1,475 1,764 797 2,212 9' 4,871 15,407 14,036 11,412 10,292 9,565 8,770 6,417 6,611 3,988 4,821 2,574 7,774 79 3,547 9,745 7.755 6,880 6,397 5,776 5,387 3,654 4.179 2,466 3,252 1,692 6,417 56 7,860 11,773 11,917 . 10,941 11,208 11,668 12,202 10,362 12,761 10,404 13,532 11,535 47,899 108 5,118 7,45 1 7.598 7,6o5 • 8,254 9.443 11,319 9,874 13,454 10,904 i5,839 12,972 62,416 i33 371 706 1,971 . 3,346 4,786 7,306 10,086 10,143 12,683 8,331 9,825 4,143 9,915 71 216 484 1,069 ~ 1,61 5 1,973 2,480 3,181 2,656 3,46s 2,1 1 5 2,709 1,247 3,58o 36 Total 38,103 [24,283 106,617 67,203 j 184,170 182,380 83,683 26,826 C— STATES. Age. iKeANE. Deaf Mute. Blind. Lepebs. Males. Females. Males. Females. Males. Females. Males. Females. 0—4 5-9 10 — 24 1 5 — 19 ■-'■'- zo — 24 25 — 29 •0-34 35 — 39 40—44 45—49 50—54 55—59 60 and over Age not returned 141 570 723 ^778 805 900 896 619 692 360 418 170 385 81 no 325 422 485 486 396 456 3l2 392 2l3 286 106 346 32 581 1,679 1,722 1,556 1,368 1,249 1,144 865 887 564 666 379 1,099 121 5io 1,224 i,o32 965 866 , 793 736 521 624 327 446 ai7 833 67 1,566 3,068 2,925 2,517 ' 2,703 2,804 2,954 2,454 3,224 2,510 3,631 2,734 11,873 125 i,25o 2,070 1,839 1,783 2,o38 1.417 2,898 2,402 3,820 2,490 4,539 3,o3i i6,583 75 55 137 314 511 783 • 1,036 1,345 1,352 1,791 1,157 1,240 544 1,216 54 69 123 236 319 304 387 497 434 549 288 359 168 443 24 Total 7,536 4,367 13,880 9,161 i 45,083 47,235 11,535 4,200 b 2 Xll XIII.— Occupation. Order of Occupation or Means of Livelihood I. Administration by State or by Local Bodies II. Defence, Military and Naval III. Service of Foreign States t - ' IV. Provision and care of Cattle V. Agricultural VI. Personal, Household, or Sanitary Services VII. Provision of E'ood and Drink VIII. Provision of Light, Firing, and Forage IX. Construction of Buildings X. Construction of Vehicles and Vessels XI. Provision of SupplercentLiry Requirements XII. Provis on of Textile Fabrics and Dress XIII. Provision of Metals and Precious Stones XIV. Provision of Glass, Pottery, and Stoneware XV. Provision of Wood, Cane, Mats, &c. XV^I. Provision of Drugs, Dyes, and Gums XVII. Provision of Leather Hides, and Horns XVIII. Commerce XIX. Transport and Storage XX. Learned and Artistic Professions XXI. Sport and Amnsements XXII. Earthwork and General Labour XXIII. Undefined and Disreputable Means of Livelihood^ XXIV. Means of Livelihood independent of Work Total Number of Persons supported by each Order. India. Provinces. 5,6oo,iS3 664,422 Soo,o3o 3,645,849 171,735,390 11,220,072 14,575,593 3,522,257 1,437.739 146,508 1,155,267 12,611,267 3,821,433 2,360,623 4,293,012 391,57s 3,285,307 4,685,579 3,952,993 5,672,191 141,180 25,468,017 1,562,981 4.773,993 States. 3,839.643 334.193 38,179 2,472,872 135,504,696 8,So5,420 12,120,669 2,887,525 1,113,633 135,627 991,334 9,655,213 2,897,046 1,669,019 3,319,170 319,981 2,224,604 3,093,056 3,242,281 4,386,725 98,48s 18,414,315 704,801 3,204,46s 287,223,431 221,172,952 1,760,510 3 3 0,229 461,851 1.172.977 36,230,694 2,714,652 2,454.924 634,732 324,106 10,881 163,933 2,956,054 924,387 691,604 973,842 71.594 i,o6o,7o3 1,592,523 710,712 1,285,466 42,695 7,053,702 858,i8o 1,569,528 66,050,479 » In this return no distinction is drawn between those who work and those whom they support by their work. The whole population depending upon the occupation is included, in order to indicate the respective sustaining power of the different orders. f Order III. is intended to refer solely to those not in the employ of the State making the return, but it is probable that in some cases the servants of the State itself have bean included. f Order XXIII. includes those not enumerated by occupation in certain wild tracts of Bajputana, Kashmir, and the Bombay States. XJU TABLE A. GENERAL STATEMEISTT. EXPLANATORY NOTE. In addition to the population shown in this table, which was dealt with in connection with the Census, the following tracts were roughly enumerated under a system of family or tribal registration during the years 1890-91, but the results being approximate only, have not been included in the Census tables. Other tracts on the Upper Burma frontier were duly enumerated, but, as in Manipur, the returns were destroyed in the subsequent disturbances. The population, however, save in Manipur, was roughly totalled at the time of Census, so the results are entered below : — Tracts. Approximate Population. 1. Upper Burma Frontier (in Bham6 and Katha) 2. British Balochistan, excluding Quetta and Railways, &c. 3. Burma Frontier returns, since destroyed Total, British 4. Sikkim 5. Shdn States 6. Rajputana (Bhils, &c.) Total, States Total excluded 42,217 145,417 74,276 261,910 30,458 372,969 204,241 607,668 869,578 2. Amongst the tracts included in this table are several which were not enumerated at the Census in full detail. The population of such tracts will be found excluded from the returns for which the information concerning it is not available, the necessary explanation being given on the title page of the table in question. b3 TABLE A. ABKA OCCUPIED HOUSES. PROVINCE. in Square Miles. TOWNS. ii i VILLAGES. TOTAL. In Towns. In Villages. I. '^. 3. 4. 5. 6. 7- I. Ajmer Meewara 2,711 4 741 101,654 25,787 75,867 ■J.. Assam 49,004 18 17,14a 1,118,885 22,399 1,096,486 3. Bengal ' i5i,543 ~ 146 227^019" 13,592,154 656,1 3o 12,936,024 4. Berab 17,718 37 5,785 591,008 79,8o5 , 5ii,2o3 S. Bombay {Presidency) 125,14.4 216 24,988 3,380,640 594,508 2,786,132 (a) Bombay ll.'IS 192 21,361 2,857,213 533,396 3.334,916 (b) Smdh 47.789 33 3,937 ( 616,020 64.804 451.216 (0) .4de«. 80 I — 7.408 7.408 - 6. Burma {Toted) 171,430 60 28,709 1,423,604 i8i,338 1,242,266 (a) tapper - 83.473 24 I0.9S7 554,472 S0.264 474.308 (b) Lower 87.9S7 36 >7.753 869,132 101.074 768,058 7. Central Provinces 86,5oi 52 34,3o3 2,158,668 162,073 1,996,59s 8. CooBG 1,583 5 492 26,806 3,174 23,632 9. Madras 141,189 X ZI4 ' 06^6,86/ 6,709,990 608,122 6,101,868 10. N.-W. Provinces io7,5o3 484 105,716 8,225,191 1,013,547 7,211,644 (a) N.-W. Provinces 83.386 399 81,437 5,944,230 822,581 5.12'.449 (b) 0»dft 24.S17 ss 34,379 2,280,961 190,766 3.090,195^ II. PanjAb 1 10,667 178 34,664 3,127,823 393,982 2,733,841 12. QUETTAH, &e. — 2 — 4,543 4,543 — l3. Aa DAMANS — — 59 2,997 — a.997 Total, Provinces 964,993 1,416 536,485 40,463,963 3,745,408 36,718,555 STATE OR AGENCY. . H. Haideeabad - 82,698 76 20,011 2,283,787 235,426 2,048,361 iS. Baroda 8,226 40 3,oo3 538,967 i23,3oo 415,667 16, Mysore 27,936 98 16,784 894,446 114,182 780,264 17, Kashmir 80,900 8 8,3io 447,993 36,252 411.741 18. Kajphtana 1 30,268 r 124 30,299 2,177,425 293,806 1,883,619 19. Central India 77,808 66 32,415 1,961,771 202,999 1,758,772 zo. Bombay States 69,045 119 i5,332 1,596,132 244,680 1,351,452 21. Madeas Statbb 9,609 14 1.293 726,966 32,826 694,140 22. Central Prot. States 29.435 6 10,401 409,096 8,761 400,33 s 23. Bengal States 35,834 2 18,804 584,912 3.5a3 581,389 24. N.-W. Prov. States 5,109 6 2,3l2 132,815 5,067 127,748 25. Panjab States 38,299 60 20,0 5 5 713,7.35 82,j65 631,570 26. ShAn States (Fort Steadman) — — 10 94 — 94 Total, States 595,167 619 179,029 12,468,139 1,382,987 11,085,152 INDIA 1,560,160 2,035 715,514 52,932,102 5,128,395 47,803,707 XV lABLE A. TABLE A. / POPULATION. Both Sexes. Males. Females. TOTAL. Urban. Rural. TOTAL. Urban. Bural. TOTAL. Urban. Rural. 8. 9- 10. II. 12. 1 3. 14. i5. 16. 542,358 ii8,63i 423,727 288,325 64,466 223,859 254,033 54,i65 199,868 5,476,833 101,074 5,374.759 2,819,575 63,495 2,756,080 2,657,258 38,579 2,618,679 71,346,987 3,443,876 67,903,111 35,563,299 1,905,224 33,658,07s 35,783,688 .,538,652 34,245,036 2,897,491 . 360,711 ^,536,780 1,491,826 187,965 i,3o3,86i 1,405,665 172,746 1,232,919 18,901,123 3,502,678 15,398,445 9,793,981 1,913,082 7,880,899 9,107,142 1,589,596 7,517,546 15,985,270 3,116,304 ia,S68,966 8,194,477 1,691,583 6.502,894 7,790,793 1,424,721 6,366,072 2,871,774 34I.29S 3,539,479 1,568,590 190,58s I,378,oo< 1,303,184 151,710 1,151,474 44,079 44,079 - 30,914 ' 30,914 — 13,185 13,165 — 7,605,560 946,649 6,658,911 3,876,301 536,797 3,339,504 3,729,259 409,852 3,319,407 2,946,933 371,404 3,575,5^9 1,414,005 186,371 1,227,634 1,632,928 185,033 1,347,895 4,658,ea7 S7S,34S 4,083,38s _ 2,462,296 3So,4j>6 2,111,870 2,196,331 224,819 1,971,5" 10,784,294 739,592 10,044,702 5,397,304 379,728 5,017,576 5,386,990 359,864 5,027,126 173,055. i5,5ii J 57.544 95,907 8,63o 87.277 77,148 6,811 70,267 35,630,440 3,406,105 32,224,335 17,619,395 1,663,790 15,955,605 18,011,045 i,74^3i5 16,268,730 46,905,085 5,314,328 41,590,757 24,303,601 2,792.279 2I,5lI,322 22,601,484 2,522,049 20,079,435 34,251,264 4.3S'.S73 29,901,681 17,812,860 3,286,339 15,526,5" 16,441,404 2,066,234 14.375.170 12,650,831 961,755 1 1,689,076 6,490,751 505,94° 5,984.811 6,166,080 455,815 5,704,265 20,S66,847 2,41 3,704 18,453,143 11,255,986 1,356,677 9,899,309 9,610,861 1,057,027 8,553,834 27,270 27,270 — 23,864 23,864 — 3,406 3,406 — 15,609 — 15,609 13,375 — .3,375 2,234 — 2,234 221,172,952 20,391,129 200,781,823 112,542,739 10,895,997 101,646,742 108,630,213 9,495,132 99,135,081 11,537,040 1,090,129 10,446,911 5,873,129 557,071 5,3i6,o58 5,663,911 533,o58 5,i3o,853 |, 2,415,396 483,5i5 1,931,881 1,252,983 248,643 1,004,340 1,162,413 234,872 9*7,541 4,943,604 626,558 4,317,046 2,483,451 311,664 2,171,787 2,460,153 314,894 ^,145,259 1 2,543,952 197,743 2,346,209 1,353,229 109,552 1,243,677 1,190,723 88,191 1, 102,532 12,016,102 1,530,087 10,486,015 6,353,488 791,514 5.56i,974 5,662,614 738,573 4,924,041 10,318,812 964,538 9.354.274 5,395,536 514,289 4,881,247 4,923,276 450,249 4,473,027 I 8,059,298 'i 1 3,700,622 1,177,422 6,881,876 4,120,125 592,865 3,527,260 3,939,173 584,557 3,354,616 175,125 3,525,497 1,853,976 88,678 1,765,298 1,846,646 86,447 1,760,199 2,160,511 38,656 2,121,855 1,089,011 20,166 1,068,845 1,071,500 1 8,490 i,oS3,oio |.. 3,296,379 16,542 3,279.837 1,673,186 9.771 i,663,4i5 1,623,193 6,771 1,616,422 '% 792,491 io3,i88 689,303 409,470 54.478 354,992 383,021 48,710 334,311 4,263,280 456,544 3,806,736 2,324,091 251,472 2,072,619 1,939,189 205,072 1,734,117 2,992 — 2,992 2,882 — 2,882 110 — no 66,050,479 6,860,047 59,190,432 34,184,557 3,550,163 30,634,394 31,865,922 3,309,884 28,556,038 1287,223,431 27,251,176 259,972,255 146,727,296 14,446,160 132,281,136 140,496,135 12,805,016 127,691,119 b 4 XTU TABLE B. VAKIATION IN POPULATIOIV SIJSTCE 1881. EXPLANATORY NOTE. In this return a distinction is drawn in the case of both Provinces and States between the tracts included in the Census of 1881 and those now brought under the operations for the first time. The total population for 1891, as here given, falls short of that returned in Table I. by 43,716 persons, belonging to tracts in Rajputana not enumerated on the former occasion. Similarly, in order to get the full total of 1881, 309,380 persons must be added, representing the Rajputana Bhil tracts not enumerated in 1891 ; and 221,070 representing the then population of Manipur, the returns for trhich at the present Census were destroyed during the disturbances of March 1891. The Bombay Provincial total for 1881 is increased by 1,908, the population of an alienated village, the return for which seems to have escaped compilation on that occasion, as it was believed to have been included in one of the Deccan States. A small transfer of 105 persons from Madras to Bengal, and another of 66,315 persons from the Orissa Kandh Mahals to the same Province, have been taken into considera- tion here. The 1881 population of the Jhansi territory, ceded by Grwalior since that year, has been deducted from Central India and added to the North-West Provinces. The Chomehla parganah of Jhalawar, which is partly under Central India, has nevertheless been included in Rajputana with the rest of the State ; but the Tonk parganahs in Central India, being wholly detached from Rajputana, are taken as part of the former agency. Similarly, in Table A., the detached parganahs of Grwalior and Indore in Meywar are included in Rajputana, though, as the 1881 population is not separately recorded, it has been omitted from Table B. as well as that for 1891. / Y 24230. c'jjprai a o CO 1^ *. «= ^- o CO 9, ^ t* T-l 00 t^ .-( 2 *. CO of SB s f: A ^ ? CO o oT CO 0\ M 4 CO 2 O rH S o jy- 00 IN- Ci « § CO !■ CO CO U3 CO CO CO at >o ft- 2 o sa + o + pip 02 W Eh O (4 ** I-H tf o w 5 n a o pq In M DO -^ n 1 m s P4 p-i ^ ij o pi Hi ^ » »^ o H o * o to CO 00 r-l 00 t^ "J" 5S <» 'B e« ' ' ^ TH ta + + 00 i> 00 » ci" o" CO 00 00 U3 CO CD CO ■* ■# 00 >o CO CD CO CO ^ o> 03 rH 00 ■* >o <5 CO 00 CD CO Cd »o o CO C3 CD 00 to kO 00 !> <=> '"i, CO Tt< CO 00 O oT $ N 00 ^ Oil C4 CO a> tH CO eo rH >o I~ 00 00 00 to a» U3 CO •H 4- + (S 00 CO 1-1 CO CO ■^ a> t- o> >o 00 00 a> CJ CO Oil ca ta o 00 co" CD >o CD o a ffT TjT CO e> — « OS CO us o ■* ™ rl o a> CO to o 00 ei" CO o 00 CO at o •* ■>* •0 w i> ■>*" 00 o 00 J> I I ^ CO CO of eo^ of 04 + * I- rH a^ m »o O CD O CO o CO 00 00 CD IJ-l -^ 00 + + + + VO O M GO ^ w^ VO s ^ r^ SO e- oo -* vO to 00 CO !> o o T-l r» o ■* CI rH- CO C« 00 00 eo t» CO (M CO 00 o o> oi" CO 03 CO eo U3 o CO r^ iH "* ■* 05 r-t (S ■* 03 o o in o to O t^ f-^ '-< s CO CD «> CO o o ■* to CO tH OS i~ •* - " CO '"' rH !>" 00 00 « Oil s -a s 01 I ■%!> a 3 1 > o O O H 02 ■ •■^ ■euoittppV P H p o ->( M Q ■< » « ^ » & M o b la H ^ « O I" a » o pq • ■< K H •< CO O « fH IB o < CO 15 CO o la |S| n P4 O H CO H ^■ 568 * : 228 202 6,2d6 921,637 1,416, 638 Buddhigt. 7,697 189,122 4 697 1 6} I i 2 24 6,888,075 3,844,569 4,043,506 322 i,o36 1,387 i,igi 196 5,768 1,290 7,095,398 I 5 29,608 |, .■(■ 1 ~ 1 V-L.(i:4. I 3 3;.--. 5,595 107 468 175 |tU -? 35,963 7,131,361 Zoroastrian 198 '79 412 74,263 72,411 I 534 318 96 9 87 781 39 246 342 168 74 357 39 76,952 i,o58 8,206 35 9 238>. 837 2,5ll I III. — Semitic. Musalm&n. 55 2 12,952 89,904 74,265 1,483,974 23,437,591 207,681 3,537,103 1,286,763 2,215,147 35,193 253,o3i 42.382 210,649 297,604 12,665 2,25o,386 6,346,651 4.72S.521 1,620,930 11,634,192 11,368 3,980 49,550,491 1, 138,666 188,740 252,973 1,793,710 991, 3Si 568,640 853,892 225,478 <.i.n, 11,87s 210,756 242,532 1,281,451 609 7,770,673 57,321,164 Christian. 2,683 16,844 j9o,829 1,359 161,770 151,001 7.764 3.005 120,768 8,786 111,982 12,970 3,392 865,528 58,441 49.129 9.312 53,587 3,008 483 1,491,662 20,429 646 38,i35 218 >i,855 5,999 8,239 714^51. 338. 1,655 ,i..A-. 77.-. .. ..,.322 _ 1 54 792,718 2,284,380 c 4 XXIV TABLE G.—emtinued. III. — Semitic — concluded. IV. — Antmtstio Belioions. v.— MmOB Province. Jew. ) J'orest Tribes, &c. Kinor Seligions. (0) Unitarians. (fi) TheistR. 1. i3. 14. iS. 16. 17- 1. AjMl^R 71 — I — — a. Assam 5 969,76s 25 — i3 3. Bengal 1.447 2,294,So6 17 — 6 4. Bbrar 2 137,108 3 — — 5. Bombay {Presidency) 12,465 2i3,6i8 27 — — (a) Bombay 9.429 135.6S3 35 - - (b) /Sj«d7j .... 310 ».93S - - - (C) ^£?ett . . - 3,Sa6 - — - - 6. Bdema 35i 168,449 18 — — (a) Upper S3 19,43s - - - (b) Lower 368 149,031 18 — — 7. Central Provinces 176 1,592,149 9 — — 8. COORG — — — — — 9. Madras 42 472,808 29 — 4 10. N.-W. Provinces 60' — 3 — — (a) North-West Provinces 33 - — - - (b) Owlh .... 35 — 3 — — II. Pan JAB »7 — 28 5 6 12. QUETTAH, &C. 23 — 2 — — 1 3. Andamans — 24 I — I Total, Provinces 14,669 5,848,427 163 5 30 STATE OR AGENCY. 14. Hyderabad 26 29,1 3o — — — iS. Baroda 36 29,854 — — — 16. Mysore 21 — I — I 17. Kashmir — — — — — 18. Kajpctana - - - i5 411,078 2 — — 19. Central India - 72 1,916,209 — — — 20. Bombay States 1,082 97.641 — — — 11. Mapras States 1,267 — Z — — 22. Central Pkovincb States — 489,572 I — — 23. Bengal States — 458,555 16 — 16 14. N.-W. Provwoe States — — — — — 25. PanjXb States . - - 6 — — — — 26. Shan States {Outposts) - — I — — — Total, States 2,525 3,432,040 22 — 17 INDIA 17,194 9,280,467 185 5 47 XXV TABLE C. — concluded. Table 0. AND Indefinite Ckeeds. (e) Deists. ' (,i6o,o8o 54,708 >,403 819,728 7i8S6 II. Pakjae ... 11,255,986 9,610,861 158,849 7,834 675,941 18,206 12. QUETIAH . . - 2.3,864 3,406 367 86 7,312 383 1 3. ANDAMAN'S 1.3,375 2,234 344 77 2,789 io5 Total, Provinces 112,295,457 108,404,653 2,593,887 162,248 9,903,664 447,924 B. state OB AGENCY. * 34. Hyderabad 5,873,129 5,663,911 76,522 3,238 343,566 ii,o63 i5. Baroda . . - 1,252,983 1,162,413 39,290 2,256 136,364 4,552 16. Mysore - - 2,483,451 2,460,153 61,076 6,415 200,495 11,499 17. Kashmir 69 70 8 6 53 5i 18. Kajputana - - - 11,081 5,635 465 123 2,951 483 19. Central India . - - 64,265 42,761 2,291 273 15,532 1,1-48 3.0. Bombay Siates - - - 4,109,546 3,93o,5i6 105,545 5,855 362,644 1 3,404 21. Madras States 1,853,976 1,846,646 82,318 16,160 360,746 48,380 22. Central Province States 1,089,011 1,071,500 4,991 171 18,710 627 23. Bengal States 1,601,590 1,557,347 15,792 5o3 72,642 2,026 24. N.-W. Province .States 409,470 383,021 1,969 38 12,211 3i8 i5. Pakjab States 2,324,091 1,939,189 13,404 375 123,236 2,000 »6. Shan States {Outposts) 2,882 no — I 1,221 17 Total, States 21,075,544 20,063,272 403,671 35,414 1,660,371 95,671 INDIA - 133,371,001 128,467,925 2,997,558 197,662 11,554,036 543,495 TABLE E. THE POPULATIOlSr BY BIRTHPLAOE. EXPLANATORY NOTE. This return was obtained from all Provinces and States, except a portion of the Madras Agency, Tracts of Rajputana, and from -the Gilgit division of Kaslimer, containing in all an included population of 350,508. , Thene were also 25,376 persons in the rest of India who failed to fill up the return correctly, so that the net population dealt with amounts to 286,847,547. The demand made was for the entry of the District and State of birth, if in India, and of the Country, if born elsewhere. In some instances a village which could not be identified was substituted for the former, whilst the required detail was omitted for the latter. Thus Roumania and Bulgaria were entered as Turkey : Europe, America, and the United Kingdom, instead of the_ country, and so on. The term " India," without fSrther detail, may be taken, as arule, to mean Hindustan, that is, the North- West Provinces and Oudh. ' d 3 XXX TABLE B. BiKTHPLACB. (a) In India : — I. Ajm^r-M^rwara ■1.. Assam and N^. Lushai 3. Bengal, 4. Berar 5. Bombay «. Sindh 7. Burma 8. Central Provinces 9. Coorg 10. Madras 11. N.-W. Provinees iz. Oudh i3. Punjab 14. Quettah i5. Aden 16. Andamans 17. Laccadives Total, Provinces 18. Hydrabad (Jieccari) 19. Baroda zo. Mysore and Bangalore 21. Kashmer and Jammu 22. Bajputana z3. Central India 24. Bombay States 25. Madras States 26. Central Province States 27. Bengal States and Sikkim 28. N.-W. Province States 29. Panjab States Jo. Manipur 3i. Shan States and Karenni 32. French Settlements 33. Portuguese Settlements Total States, &c. 34. India, unspecified Total, India TOTAL. Males. Females. 490,221 5,017,019 71,838,732 2,561,430 15,986,829 2,692,973 7,282,348 11,313,422 122,772 35,646,068 34,768,181 12,705,441 ' 20,608,421 3 11,713 1,909 385 221,046,867 11,539,750 2,355,870 4,889,617 2,548,726 12,478,431 10,146,485 7,950,357 3,629,824 1,579,582 2,878,968 788,135 4,206,065 5,829 8,160 23,113 54,333 65,083,245 55,798 286,185,910 259,007 2,552,409 35,763,212 1,299,530 8,i3o,65o 1,461,150 3,609,127 5,639,959 62,836 17,679,522 18,197,223 6,572,501 ii,i52,53i 2 5,810 1,523 364 112,387,356 5,854,734 1,22I,5Z4 2,445,278 i,35o,722 6,626,158 5,283,139 4,089,^04 1,820,108 796,879 1,454,697 407,433 2,278,294 3,5o3 5,o3i 9.325 34,557 33,681,186 31,490 146,100,032 231,214 2,(64,610 36,075,520 1,261,900 7.855,179 1,231,823 3,673,221 5,673,463 59,936 17,966,546 16,570,958 6,132,940 9,455,890 I 5,903 386 21 108,659,511 5,6X5,016 1,134,346 2,444,339 1,198,004 5,852,273 5,863,346 3,860,553. 1,809,71 6. 782,703. 1,424,271 380,702 1.927.771 2,326 3,119 13,788 19,776 31,402,059 24,308 140,085,878 XXXI Table E. TABLE E.—conUnued. (5) COUNTKIES ADJACENT TO INDIA 1 35. Beluchistan 35, Afghanistan 37. Yaghestan 38. Turkestan 39. Tibet 40. Nipal 41. Bhutan 42. Oejlon 43. Straits Settlements, &o. 44. Slam Total, adjacent Countries (e) ((THER Asiatic Countkies :— 45. China - - " 46. Mongolia {Tartary') 47. Japan 48. Hong Kong 49. Macao, Manilla, &c. 50. Java and Sumatra 51. Persia 52. Levant and Armenia 53. Arabia and Baghdad 54. Russian Turkestan 55. Asia, unspecified Total, other Asiatic Countries (d) EuEOPE :— 56. Kngland and Wales 57. Scotland S'S. Ireland 59. Channel Islands 60. United Kingdom, unspecified Total, United Kingdom 61. Gibraltar 6z. Malta 63. Cyprus (4. France 65. Belgium f6. Holland . - - 478,656 25,563 172 125 8 42 ]00 4,411 1,874 28,092 52 80 60,519 79,172 8,580 12,706 141 100,551 104 199 1 1,258 214 123 278,332 22,470 122 5o 6 33 59 3,35i 1,337 22,829 45 +7 50,349 70,li3 7,309 10,587 io5 I 88,115 73 i37 800 128 ii5 200,324 ■3,093 So ',S 2 9 41 .1 ,06.0 53 5,263 7 33 10,170 9.059 1,121 .1,119 36 3.S04 Lower . . - - 4,658,627 3.17s 1.773 4.374 2,96c, Central Provinces 10,784,294 1,758 5,463 18,567 5,355 GooKG ;- 173,055 44 126 86 23 Mat>ras ;- 35,630,440 7,687 26,983 36,424 12,617 N.-W. Provinces 46,905,085 5,640 33,147 111,039 17,071 N.-W. Provincet 34,254,254 4,103 33,228 7S,3i6 11,768 Oudh . . . ■ 12,650,831 l,S3S 9.919 35,733 S.303 Panjab 20,866,847 6,3i2 20,240 73,385 4.351 QUETTAH - 27,270 4 9 6 2 Andamans 15,609 — I I I Total, Provinces - 221,131,362 62,376 173,820 366,550 110,509 b. state oe agency. i Haiderabad 11,537,040 1,584 4.419 io,632 1.977 Baroda 2,415,896 845 918 4.751 569 Mtsore . . - 4.943,604 1,07 s 3,418 5.194 g02 Kashm:^e - — — — — — Rajputana 11,990,504 3,097 *2 38,280 1,708 Central India (^Railways and Canton- ments). 99,686 *69 *49 *43. go Bombay States 8,040,062 1.994 4.697 13,028 ^,554 Madras States 3,7p0,622 677 1.557 2,309 '.439 Central Province States 2,160,511 276 977 1.903 1.2S9 Bengal States 3,296,379 i,oS3 2,482 2,934 2,048 N.-W. Province States - 792,491 i57 393 1,024 379 Panjab States 4,263,280 1,076 4.129 11,832 1,920 Shan States ( Outposts) - 2,992 — — — ^_ Total, States 63,242,567 11,903 •23,041 92,318 16,736 INDIA 274,373,929 74,289 *196,861 458,868 126,244 * Se« the Explftnatory Note XXXV Table F. TABLE F. MALES. FEMALES. Total Males. N«mber of Insane. Deaf Mutes. 9- Blind. Lepers, II. Total Females. Number of Insane. i3. Deaf Mutes. H- Blind. iS. Lepers. 1 6. 288,325 2,798,908 35,563,299 1,491,826 9,793,981 8,194,477 1,368,590 30,914 3,876,301 1,414,005 2,462,296 5,397,304 95,907 17,619,395 24,303,601 17,812,850 6,490,7.')1 11,255,986 23,864 13,375 112,522,072 5,873,129 1,252,983 2,483,451 6,340,443 59,139 4,109,546 1,853,976 1,089,011 1,673,186 409,470 2,324,091 2,882 27,471,307 139,993,379 64 J,737 14,357 280 4,1 3o 2.J1S 1,608 3,79s 2,042 1,127 z5 4,463 3,886 2,830 i,oS6 4,235 4 38,103 i,o36 535 618 2,024 ♦40 1,224 402 167 663 , III 716 7,536 45,639 Ii3 2,683 41,373 317 7.149 S.13S 3,008 9 2,1 5o i,"5 1,03s 3,i59 77 1 5,274 21,384 14,981 6,403 12,928 9 I 106,617 2,729 568 1,908 *26 i,8oi 875 570 1,572 24S 2,586 *13,880 *120,497 523 3,o3i 35,614 3,38o 14,640 11,340 3,357 23 6,681 4 All 2,204 8.444 47 I7.7>5 56,o58 38,466 38,o34 3 184,170 - 5,892 2,017 2,644 17,276 *245 6,104 1,271 923 1,463 543 6,705 '45,083 229,253 20 5,128 32,957 2,886 7,684 7,S58 I 4,543 2,262 2,281 3,575 12 9.439 14,114 9.574 4,S40 3,322 2 I 83,683 2,261 397 536 1.314 *59 1,911 1,0 1_3 799 1.47 > 3l2 1,462 11,535 95,218 254,033 23 2,636,335 1,285 35,783,688 9,808 1,405,665 198 9,107,142 2,166 7,790.793 1,328 1,303,184 83i! 13,165 3 3,729,259 1,532,928 2,196,331 5,386,990 77,148 18,011,045 22,601,484 16,441,404 6,160,080 9,610,861 3,406 2,234 108,609,290 5,663,911 1,162,413 2,460,153 5,650,061 1,073 40,547 "29 3,930,516 770 1,846,646 275 1,071,500 109 1,623,193 383,021 1,939,189 110 25,771,260 134,380,550 3,088 i,9SS 1,133 63 1 19 3,224 1.754 1,272 4S2 2,077 24,273 548 3io 457 390 46 360 4,367 28,650 62 1,998 25,564 2l3 4.47 S 3,482 986 7 1.754 1,016 738 2,304 49 11,709 11,763 8,247 3,Si6 7,3i2 67,203 1,690 35o i,5io *23 682 407 910 148 1,543 *9,161 *7 6,364 53o 7 2,801 1,599 34,442 11,029 3,388 812 13,459 2,5o3 io,5S9 2,419 2,878 84 8,53 6,383 2,170 10,123 39 18,709 54,981 36,850 18,13: 35,35i 3 t 182,380 4.740 2.734 2,S5o 21,004 '186 6,924 i,o38 980 '.47> 481 5,127 47,235 1,921 1,242 679 1,780 II 3,178 2,957 2,194 763 1,029 26,826 716 172 266 394 21 643 426 460 577 67 458 4,200 229.615 31,026 preceding this Table. e 2 xxxvu TABLE G. THE CHRISTIAIV POPULATION BY RACE AND DENOMINATION. EXPLANATORY NOTE. In tbe instructions accompanying the sctedule all Christians were asked to return their denomination, whilst as regards other religions this detail was left optional with the local administration. In order to secure as complete a return of Christian sab-divisions as possible the various missionary bodies were addressed with a view to the issue of instructions from their different centres as to the correct entry to be made by their congregations. In most cases the entry was printed in the vernacular on a card which was to be shown to and copied by the enumerator, but the return makes it plain that this precaution was to a considerable extent disregarded. n XXXVUl TABLE G. Total Returned. Distribution by Eace. Desomikation. European. Eurasian. Native and African. I. a. ■ 3. 4- 5. Abyssinian 41 4X Armenian 817 366 247 204 Sjrian {Jacobite) 200,467 iS 3 200,449 Boman 1,315,263 35,64s 36,089 1,243,529 Greek 400 36o 28 12 Church of England 29.5,016 101,871 29,116 164,028 „ India -" 6 — .._ 6 Anglican Church 3 2 1 — " Episcopalian " 1,919 azo 20S 1.494 •• Evangelical " a 5 — 6 Evangelic Union 3 2 ~ I " Evangelist Church " - 199 3i . 167 "Keformed" 1 I — — Eondon Mission 20 2 10 3 St. James' Church .... 6 — — 6 St Paul's „ . ■ 2 I 1 — Church of Ireland 55 39 -'^--y 16 ,, Scotland 4,176 1.7S1 1,293 • I,l32 „ America - 1,014 11 — 1,002 Free Church 145 — — 14S of Scotland 1,010 3o5 44 661 Presbyterian 40,407 g,4i8 1,074 30,91 5 United Presbyterian 82 gz — Keformed Presbyterian 53 — — 53 American „ 454 24 29 401 Irish 2 2 — — Irish Mission 22 — — 22 Independent 191 45 I 145 Congregationalist 7,703 zzg 129 7,346 Methodist - . . 9,957 i,o36 900 8,021 Methodist Episcopalian 14,603 449 642 13,412 Primitive Methodist 4 4 — Wesleyan 6,388 3,129 684 2.57S Wesieyan Methodist 1,269 743 122 404 Baptist ... 191,746 2,907 2,352 186,487 Anabaptist - . . - . 2 — 2 — XSXIX Table G. TABLE G. Denomination. Total Keturned. Distribution by Bace. European. Eurasian. Native and African. Plymouth Brethren " Open " Brethren Bible Christian " Nonconformist " " Dissenter " Puritan Society of Friends Quaker " Protestant " Calvinist Welsh Calvinist Lutheran German Church German Mission Swedish Church Beformed Dutch Zwinglian - Moravian Catholic Apostolic - Salvationist " Unseetarian "♦ « Heathen " (Convert) Swedenborgian - New Jerusalem - Unitarian (sic') - - " Sarweswara Mathd (Unitarian) Theist (sic) Agnostic (sic) Denomination not returned* Total Christians - 32 1 2 127 428 2 4 108 60,713 3 65,376 6 1 1 4 1 2 6 1,286 2,343 5 1 202 3 1 2 60,352 2,284,380 z6 Z04 4 28 6,i38 3 6 939 I 4 I 5 114 927 5 I 69 I 1,703 168,000 I I 38 14Z 5,zS2 194 I 34 347 758 79,790 63 82 2 80 49,223 64,243 5 i,i38 1,069 2 85 3 57,891 S,036,590 . , , . „ »!.„ „~ooc vtivrp the column for denomination was left blank, * Th« entry of " Unseetarian" is here distinguished from the easfsj^here the column Since the ?orK an assertion, the latter may be due to neglect or ignorance. e 4 iN^^ *A, SM6*si'*jt.; f-:^'^^. m. ^■■|^ %f.i t^*^ Cag JJ ^ &>:^^J- ^^i ^•i«^. Lit*;; :,3Si^i!C £Sa^;^^ES. ^-iiUiSi