m ',» Jlr^!»" (. ^1 • ^^m^M^^^'^''''-j!'Wf^^^^ m^. 1' v't.i'i py'-irn-^'^v'-i- W^KBKta'i^'V'-i' ■■■■ , Albert K. Mann Libkary Cornell University N.V.S. m. LIBRARY CORNELL UNIVERSITY. THE THE GIFT OF ROSWELL p. FLOWER FOR THE USE OF THE N. Y. STATE VETERINARY COLLEGE. 1897 ILL UNIVERSITY LIBRARY 3 1924 094 811 357 Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924094811357 Report No. 64, U, S. Dept Agr. Plate I. q: -^ LU SJ I- ^ < o > ic 2 a <3 i^ I- '- o CO o CO o = i ■- iti ■» > S. 30. U. S. DEPARTMENT OF AGRICULTURE. Report IS"©. 6-4. lIVISIi 1899. BY MILTON ^A'■HITNEY, CHIEF OF DIVISION OF SOILS, \? WITH ACCOMPANYING PAPERS BY THOMAS H. MEANS, CLARENCE W. DORSET, FRANK D. GARDNER, FRANK K. CAMERON, LYMAN J. BRIGGS. PEINTBD BT ACT? OF CONGEESS. WASHINGTON: GOVERNMENT PRINTING OFFICE. 1900. /iAWtO s i-v 14 P ^OQ To the Senate and House of Representatives: I transmit Lerewitli for tlie information of the Congress a communi- cation from the Secretary of Agriculture, covering extensive field oper ations, consisting of soil surveys over various areas, aggregating 720,000 acres. In the opinion of the Secretary of Agriculture, this is the most important work of its kind ever undertaken, William McKinley, Executive Mansion, February 5, 1900. ij LETTER OF TRANSMITTAL cr. S. Department op AaBicuLTURB, Office of the Secretary, Washington, D, 0., February 1, 1900. Mr. President : I have the honor to transmit a report of this Depart- ment upon the field operations of the Division of Soils during the year 1899. It covers not only the most important work of this division, but, is, in my opinion, regarded in the light of the possible results of the infor- mation which has been secured thereby, the most important work of this character ever undertaken in any country. It consistsof an introduction by Prof. Milton Whitney, the chief of the division, and special reports on the various lines of field work by the scientific assistants of that division, in cooperation in some instances with educational institutions in various parts of the country. The report covers, all together, a soil survey of an area of not less than 720,000 acres, scattered throughout the following regions: l!few Mexico, Utah, Colorado, and Connecticut. These accounts of soil surveys are followed by some valuable discussions on "the application of the tlieory of solutions to the study of soils," "some necessary modifications in method of mechanical analysis as applied to alkali soils," and " salts as influencing the rate of evaporation from soils." It need hardly be said that it was found quite impossible to restrict this report within the limits of a hundred octavo pages, and conse. qnently, under the operations of section 80, chapter 23, Volume XXVIII, of the Statutes at Large, 1895, the Secretary of Agriculture has no authority to secure the printing of more than 1,000 copies, a number which would be entirely inadequate to supply our exchanges, oar agri- cultural colleges and experiment stations, and the divisional list of the Division of Soils, to say nothing of any miscellaneous demand. In this case, moreover, the Department, in my opinion, is bound to supply copies of the report to the very large number of persons who have rendered efQcient aid to our surveyors in their field work. Under these circumstances and in view of the great importance of the subject covered, 1 have the honor to recommend that it be transmitted to Congress, together with the maps, illustrations, and diagrams accom- panying, to be printed by order of that body, and I further recommend that not less than 10,000 copies be printed for the use of this Depart- ment, in addition to such number as Congress may order for the use of its members. I have the honor to remain, Mr. President, very respectfully, James Wilson, Secretary. 3 LETTER OF SUBMIHAL tr. S. Department op Agkioxjltuee, Division of Soils, Washington, D. C, January 29, 1900. Sir : I have the honor to present herewith a report on the field opera- tions of the Division of Soils for 1899. This is, without doubt, the most important work that has been undertaken by the division since its organization. The report is illustrated by a number of photographs and diagrams showing the character of the country and the specific character- istics of the soil, and by accompanying maps showing the distribution of the soils, alkali, and seepage waters. That portion of the report pre- pared by Dr. Cameron relates to work undertaken in cooperation with the Division of Chemistry, and has been read and approved by Dr. "Wiley. The soil work is of great interest and is very widespread in its application. A small edition would be entirely inadequate to supply the exchange list of the Department and the ordinary mailing list of this division, besides the very large number of persons who, having aided and cooperated with the division iu its work, are entitled to a copy when published. I therefore have the honor to recommend that not less than 10,000 copies of this report be printed for the use of this Department. Eespectfully, Milton Whitney. Chief of Division. Hon. James Wilson, Secretary of Agriculture. CONTENTS. Field operations of the. Division of Soils in 1899, by Milton Whitney, Chief of Division ; Progress and cost of the soil survey Cooperation with State organizations Operations in the Pecos Valley, New Mexico Salt Lake Valley The Connecticut Valley Cecil County, Md Laboratory work Field methods for a soil survey A soil survey in the Pecos Valley, New Mexico, by Thomas H. Means and Frank D. Gardner Introduction The climate of the Pecos Valley History of irrigation in the Pecos Valley The irrigation systems of the Pecos Valley The Eoswell district The geology of the Eoswell area. Soils The Pecos sandy loam The Roswell sandy loam Eoswell loam The Hondo meadows Water supply Underground water The alkali of the soils The problems of the Eoswell area The Hagerman area Soils The water supply Underground water Alkali of the soils Problems of the district Carlsbad area Soils The"Peco8 sands The Pecos sandy loam Pecos conglomerate soil Gypsum loam or "yeso" The water supply of the Carlsbad district Underground water Alkali in soils Problems of the area - Barstow area, Texas Summary 7 8 CONTENTS. Page. A soil survey in Salt Lake A'alley, Utah, by Frank D. Gardner and John Stewart: Physiography 77 Climate 83 History of irrigation 85 Soils 88 Jordan sandy loam 89 Bingham gravelly loam .- 92 Jordan loam Si Jordan clay and clay loam 97 Jordan meadows 100 Jordan sand 100 Bingham stony loam 101 Salt Lake sand 101 Hardpan 102 Water supply 106 Application of water 109 Underground water 110 Alkali in soils Ill Black alkali 112 Uuderdrainage and the reclamation of waste lands 113 A reconnoissance in Sanpete, Cache, and Utah counties, Utah, by Thomas H. Means : Geography and topography 115 Soils 116 Alkali in soils 117 Summary 119 A reconnoissance in the Cache a la Poudre Valley, Colorado, by Thomas H. Means : Soils and alkali 121 Summary 124 A soil survey in the Connecticut Valley, by Clarence W. Dorsey and J. A. Bonsteel : Introduction 125 Topography of the valley 126 Geology 126 Climate 126 Tobacco 127 Soils 129 Triassio stony loam 130 Holy oke stony loam 131 Windsor sand 132 Hartford sandy loam 133 Podunk fine sandy loam 134 Connecticut meadows 135 Enfield sandy loam 137 Suffield clay 138 Elmwood loam 138 Connecticut swamp 139 Comparison of the texture of the soils 140 Application of the theory of solutions to the study of soils, by Frank K. Cam- eron: Introduction 141 Nature of solutions 142 Laws governing solutes 143 CONTENTS. Application of the theory of solutions to the study of soils, hy Frank K. Cam- eron — Continued. Dissociation of electrolytes 144 Reversible reactions and the Mass law 146 Heterogeneous equilibria 147 Carbonates and lime in Great Salt Lake 149 Two or more solutes with no commonion 150 Hydrolysis 151 Hilgard on the r61e of carbon dioxide 152 Gypsum and ammonium salts 155 Amelioration of black alkali 156 Analytical problems 157 Field method for estimating sulphates, chlorids, and carbonates 161 Estimation of carbonates and bicarbonates ; a field method 162 Inversion of sodium bicarbonates and sodium hisilicates 167 Genesis of hardpan 168 Ther61eof fertilizers 170 Summary 171 Some necessary modifications in method of mechanical analysis as applied to alkali soils, by Lyman J. Briggs : Introduction 173 Mechanical analysis of soils subject to excessive disintegration 173 The centrifugal method 174 Treatment of separations after ignition 178 Determination of water soluble material 180 Salts as influencing the rate of evaporation from soils, by Lyman J. Briggs : Introduction 184 Sodium chlorid 187 Sodium sulphate and sodium carbonate 191 Conclusions 198 ILLUSTRATIONS. PLATES. Page. Pl. I. Conglomerate soil on Southern Canal near Carlsbad, N. Mex., where much -water is lost by seepage Frontispiece. II. Old apple orchard, 5 miles southeast of Eoswell, N. Mex 48 III. Sand dunes, with characteristic vegetation of mesquite and canaigre, near Carlsbad, N. Mex 62 IV. Conglomerate in Pecos River bed 66 ^- -^ gypsum soil without the usual covering of loam, near Florence, N. Mex 68 VI. A lateral ditch cutting down in to the gypsum soil 70 VII. Seepage stream from the gypsum area flowing about 2 ou. ft. per sec. 72 VIII. Alkali flat caused by seepage and subirrigation of the gypsum land. 74 IX. Alkali flat formed by seepage from conglomerate 74 X. Lake Bonneville marks on mountains 3 or 4 miles east of Garfield Beach. The upper bench line is the Bonneville shore line, and the lower well-defined one is the Provo shore line 80 XI. Natural vegetation grease wood and salt bushes on Jordan sandy loam . 90 XII. Bingham gravelly loam — gravel 1 inch in diameter 92 XIII. Bingham stony loam, looking toward the lake shore 100 XIV. First stage of the formation of hardpan on the shores of the lake, with decomposing algse being incrusted with carbonate of lime. .. 102 XV. Advanced stage of formation of hardpan on shore of lake 104 XVI. North Point and Consolidated Canal, showing seepage along sides of canal 108 XVII. Last stage of vegetation with accumulation of alkali 110 XVIII. Alkali flat too strong for salt bushes 110 XIX. Alkali flat, final stage, with no vegetation 110 XX. Instruments for salt determination on right of auger and for sodium carbonate and chlorid determination on left of auger 112 XXI. Escarpment between Podunk fine sandy loam and the Connecticut Meadows 126 XXII. Triassic stony loam 130 XXIII. A drumlin or hogback of the Holyoke stony loam 130 XXIV. Windsor sand, showing old corn rows covered with forest growth . . 132 XXV. Podunk broad-leaf tobacco area 134 XXVL The Hartford sandy loam 134 XXVII. A typical tobacco bam of the Connecticut Valley, 180 feet long 136 XXVIII. Centrifugal apparatus for soil analysis 174 XXIX. Evaporation cylinders 196 FIGURES. Fig. 1. Sections in Pecos Valley, New Mexico, through Eoswell; Carlsbad; Otis ; and Florence 51 Fig. 2. Diagram showing salt content to depth of 6 feet: a. in Pecos sand; b. in Pecos sandy loam 64 11 12 ILLUSTKATIONS. Page, Fig. 3. Diagram of orchard allowing depth to standing water 66 4. Diagram of oicliard showinj; soluble salt content of surface foot 67 5. Sketch map of Pecos A'alley, showing increase in salt content of water as it flows down the valley 71 6. Sketch map of Bonneville Basin, showing ancient lake and present lakes 79 7. Diagram showing mean monthly fluctuatiou in water level of Great Salt Lake and the rainfall during same period 81 8. Diagram showing mean annual fluctuations in water-level of Great SaltLakeand the rainfall during the same period 82 9. Sections in Salt Lake Valley along lines marked on sketch map 93 10. Sketch map of western part of Salt Lake Valley, showing canals, hardpan areas, and parts of salt in 100,000 parts of river and irriga- tion waters 103 11. Apparatus for water pressure in mechanical analysis 177 12. Diagram showing evaporation from soil in trays, moistened with water and NaCl solution 188 13. Diagram showing evaporation from Petri dishes, distilled water and normal sodium chlorid solution 189 14. Diagram showing evaporation from cylinders, moistened with distilled water and sodium chlorid solution 191 15. Diagram showing evaporation from soil in trays, moistened with dis- tilled water and sodium sulphate solution 192 16. Diagram showing evaporation from soil in trays, moistened with dis- tilled water and normal sodium carbonate solution 193 17. Diagram showing evaporation from Petri dishes, normal solutions 195 18. Diagram showing evaporation from Petri dishes, saturated solutions . 196 19. Diagram showing evaporation from cylinders of soil, moistened with distilled water and with normal sodium carbonate and sodium sul- fate solutions 197 [In pocket on last cover.] Map 1. Soil map, Eoswell sheet. New Mexico. 2. Alkali map, Eoswell sheet, New Mexico. 3. Underground water map, Eoswell sheet. New Mexico. 4. Soil map, Carlsbad sheet, New Mexico. 5. Alkali map, Carlsbad sheet, New Mexico. 6. Underground water map, Carlsbad sheet, New Mexico. 7. Soil map. Salt Lake sheet, Utah. 8. Alkali map. Salt Lake sheet, Utah. 9. Black alkali map, Salt Lake sheet, Utah. 10. Underground water map, Salt Lake sheet, Utah. 11. Soil map, Hartford sheet, Connecticut and Massachusetts. FIELD OPERATIONS OF THE DIVISION OF SOILS IN 1899. By MILTON WHITNEY, Chief of Division. PROGRESS AND COST OF THE SOIL SURVEY. During the season of 1899 three well-organized parties were in the field for from six to eight months each, equipped according to the most modern methods for surveying, investigation, and mapping the soils of several important agricultural districts. Eleven hundred and twenty- five square miles, or about 720,000 acres, have beeu surveyed and mapped on a scale of 1 inch to the mile. The following table shows the areas which have been surveyed, the rate per day, and the actual cost of the field work per square mile : Eate per day. Cost per square mile. Cecil County, Md CoDnectiCDt Valley Salt Lake Valley, Utah Pecos Valley, New Mexico. Sqitare miles. 375 400 250 100 Square miles. 5 4.5 2 1.1 $1.70 2.00 a6.no 12.00 Total 1,125 a The actual cost of the work to the Division of Soils was $4 per square mile, the Utah Experiment station cooperating and paying the equivalent of $2 per square mile. The difference in the cost of the work is due to the more complex problems presented in the arid regions of the West, as will be explained, as well as to the greater distance from Washington and the consequent greater cost of transportation. The cost per square mile given in the table includes the salaries of the men while actually in the field, together with their transportation and subsistence expenses. Taking into account the annual salaries of the men assigned to the soil survey, together with all expenses directly chargeable to the work in the field, the amount expended during the calendar year 1899 for such work amounted to 27.8 per cent of the total appropriation for the Division of Soils. The laboratories, contributing largely to the efflciency and support of the soil surveys, consumed 22.5 per cent of the total appro- priation; the administrative expenses — including rent of building, express, gas, and the fitting up of an additional laboratory room — amounted to 40.5 per cent of the total appropriation. 13 14 FIELD OPERATIONS OP THE DIVISION OF SOILS. For administrative purposes in the Division of Soils the United States has been divided into the eastern and western districts, with the Mississippi Eiver as the dividing line. Mr. C. W. Dorsey has gen- eral charge of the soil survey of the eastern district, and Mr. Thos. H. Means of the western district. In the western district the work has so far been confined to irrigation districts, and has included, besides the mapping of the soils, the investigation and mapping of the alkali conditions and of the depth to standing water. Such maps, showing the exact distribution and condition of the soil, serve as a basis for the improvement or reclamation of alkali lands by underdrainage or by other means for removing the excessive accumulation of alkali and seepage waters. They also show the source of possible future damage, and by their aid suitable preventive measures can be taken to provide against such losses as have occurred in the past. In localities where sodium carbonate or black alkali prevails a sep- arate map is made, showing the distribution and relative intensity of this substance. The areas which have been surveyed this year are not at all simple in their conditions, and represent perhaps as great a variety of soils and soil conditions as will ever be encountered in a season's work. As a rule, the soils of the West are more uniform over small areas than they are in the Atlantic Coast States, but on account of the greater number of conditions to be taken under consideration and investigated, in the surveying and mapping, and the necessarily more complex field methods, the work is much slower. The expense of the work in Cecil County, Md., and in the Connecti- cut Valley averages less than $2 per square mile, including the salaries of the men while in the field and the transportation and subsistence expenses. This is considered very satisfactory. The parties consisted in each case of two men working together, each party having a horse and wagon. The relatively small area per day covered in the Pecos Valley, New Mexico, and the apparently excessive cost ($12 per square mile) is accounted for by the newness of the work, the fact that two experienced men were assigned to this work — -for their mutual training and for the iinal perfecting of the methods — and to the further fact that consider- able detail work was done in the study of the conditions and methods to be used under such conditions — necessary work which will not have to be done again. The expenses of the work in Salt Lake Valley, Utah ($6 per square mile), are considered fairly satisfactory, although it will doubtless be possible to materially increase the rate per day, and thus lessen the cost per square mile as the men become better acquainted with the methods of surveying and with the causes and conditions of the alkali soils. With the experience of this season such an area as this could probably be surveyed now at a cost of about $4 per square mile. COOPERATION WITH STATE 0KGANI2ATI0NS 15 It must be remembered that the estimation of the cost of such work in new localities is a mere approximation, as the expense depends largely upon the complexity of the conditions in the locality to be surveyed. In each of these districts in the West the field parties consisted of two men, with a horse and wagon for carrying the field equipment, without a camping outfit, as the work done was in rather closely settled districts. It has been the intention to have an experienced man in charge of each field party, with a student assistant or an intelligent laborer to assist in the work. So far as possible advantage is taken of these oppor- tunities to train men for the more responsible positions devolving upon directors of field parties. Such educational work is a very essential and important part of the work of the division, as it is impossible to obtain men already trained and fitted for this field work. COOPERATION WITH STATE ORGANIZATIONS. Wherever possible it has been the policy of the division to work in close cooperation with the State experiment stations, the State geo- logical surveys, boards of agriculture, or other local institutions. This cooperation, so far as it has been carried, has proved mutually satis- factory and beneficial to the local institutions and to this division. In a general way it relieves the local institutions of all responsibility of the direction of the work and the preparation of the material, while the expense to them is very much less in every case than if they had to employ competent soil experts, even if such experts could be obtained, which is questionable. On the other hand, this cooperation enables the Division of Soils to very materially widen its sphere of field work, and gives the division a desirable connection with local institutions, in touch with the people actually interested in the work. Owing to the difference in the character of the work in the Eastern and Western districts the basis of the cooperation with State institu- tions has been essentially different. The arrangements under which the work has been done during the present season may be briefly stated. The soil survey of the Connecticut Valley was carried on by the Division of Soils, independently of any other organization, under the authority of Congress to investigate and map the tobacco soils of the United States. This work was in charge of Mr. 0. W. Dorsey, with Mr. J. A. Bonsteel, of the Johns Ilopkins University, as temporary assistant. A detailed soil map, with accompanying text describing the different soils, is given in the latter part of this bulletin. Under authority from the Secretary of Agriculture an agreement was entered into with the Maryland geological survey, Dr. William B. Clark, director, to survey and map the soils of Maryland at the rate of one county each season, which is the present rate at which the topographic base maps and the geological data are being prepared. Under the terms of this agreement the Division of Soils assumes the 16 FIELD OPERATIONS OF THE DIVISION OF SOILS. entire responsibility for the details of the soil survey, together with the preparation of the manuscript map, with a brief description of the soils. This material is turned over to the State survey, the Department reserving the prior right of publication, to be published with such additional data as they may wish to include in their final report. The publication of the map by the Division of Soils, with brief notes as to the soil classification, is not expected to conflict in any way with the more elaborate State publication for local distribution. Under tbe terms of the agreement the Maryland Geological Survey pays the trav- eling and subsistence expenses of the field party. This amounted in the present case to about $400, or approximately $1 per square mile. The area selected by the director of the State survey was Cecil County. While the soil survey has been completed and the manuscript map pre- pared, the publication will be delayed a year, awaiting the engraving and publication of the topographic base sheet, which has just been com- pleted in manuscript. The soil survey of the Pecos Valley, New Mexico, was carried on by Thomas H. Means and Prank D. Gardner, of this division. The entire expense was paid by the Division of Soils, and there was no official cooperation with any other organization. The New Mexico Experi- ment Station, however, authorized Prof. J. D. Tinsley to accompany the field party for the purpose of studying our methods. Professor Tinsley spent about a month helloing with the work in the district around Eoswell. The work in Utah was done in cooperation with the Utah Experiment Station, Prof. Luther Foster, director. The terms of the cooperation were that the Division of Soils would send an expert to take charge of a field party, pay his salary, traveling and subsistence expenses, and be responsible for the preparation of maps and reports which would treat fully of the conditions of the area. The Utah Experiment Station would furnish an assistant and pay his salary and traveling and sub- sistence expenses while in the field. The expense to the experiment station amounted to about $2 per square mile. In this alkali work, unlike the soil work in the eastern district, the reports have to treat very fully of all the conditions, and the same report will therefore serve for and be credited to the Division of Soils and to the Utah Experiment Station. The director of the Utah Experiment Station designated Mr. John Stewart as an assistant in the Utah work, Mr. Frank D. Gardner hav- ing direct charge of the field party. The results of the field work are briefly referred to in this paper and are set forth in detail in the accom- panying papers and maps. In the progress of the field work this year, and particularly in the review of the work in the preparation of the reports and maps, many valuable modifications have been suggested in our methods which it is believed will greatly simplify the work and increase the efficiency OPERATIONS IN THE PECOS VALLEY, NEW MEXICO. 17 of the field parties. One feature which has been brought out clearly is the fact that the field . methods are now practically independent of the laboratory methods. The soils and alkali salts are classified and mapped in the field. The maps themselves are actually prepared in the field in such form that they can be turned over to the lithographer after suitable titles and legends are prepared. This refers to the soil, the alkali, the black alkali, and the underground water maps. Much of the laborious. work of calculating the actual salt content has been saved, as limiting values only are shown upon the map, and ifc is hardly con- sidered advisable to publish the figures of actual salt determinations in tables which few, if any, would read. The laboratory work connected with the survey has likewise been reduced to a minimum. The chemical and physical laboratories are used now simply to explain the character of the soils which have formed the basis of the classification. For example, the physical differences between the Hartford sandy loam and the Connecticut meadows are clearly brought out in the mechanical analyses. On the other hand, the differences between the Pecos gypsum soil and the Pecos sandy loam are apparent not only from the mechanical analysis, but also through the chemical analysis. These conditions, however, are so marked in each case and so apparent that the soil areas can be accu- rately mapped, even without the results of the physical and chemical examinations. This matter of the field methods will be referred to more at length on a subsequent page. The alkali and underground water maps show the conditions at the time of the survey. At different seasons of the year and especially under different weather conditions the soil conditions maybe somewhat different} but as the examinations are usually made during the grow- ing season the conditions shown may be taken to represent fairly well the actual conditions of growth. OPERATIONS IN THE PECOS VALLEY, NEW MEXICO. Reports have come in to this Department from time to time for several years from the Pecos Valley of root-rot disease affecting fruit trees, grapevines, sugar beets, and alfalfa — the trees and vines much more seriously than the other crops. Truck crops seem to have largely failed in the valley, while the yield from alfalfa, which seems to be the crop best adapted to the conditions there, is only about 5 tons to the acre in 4 cuttings. In Montana it is from 6 to 7 tons in 3 cuttings. In view of these serious reports, the importance of the locality, and the expressed willingness of the people to benefit by the results of the investigations, it was decided to send an expedition to investigate and map the soils and soil conditions of a portion of the valley. Accord- ingly, Thomas H. Means and Frank D. Gardner, of this division, were sent out and spent two months at Carlsbad and one month at Roswell. I personally inspected the work during the progress of the investiga- tions. The detailed account of this work is given in an accompanying H. Doc. 399 2 18 FIELD OPERATIONS OF THE DIVISION OF SOILS. l^aper, and is illustrated by a series of maps, diagrams, and i^hoto- graphs. It was found in the Carlsbad area that fruit, truck, and grapes were almost entire failures; that the sugar beets and alfalfa gave relatively small yields and were affected with root-rot over certain areas. There are many abandoned farms and large areas over which the conditions generally are not satisfactory. One reason that fruit has not been successful is the liability of the occurrence of late spring frosts which destroy the buds after they have started. The root-rot disease, which has been so destructive in the valley, is unquestionably largely due to the condition of the irriga- tion water during part of the year and over some sections of the valley. The salt content of the water is so high that it is near the limit of endurance for crops. The occurrence of gypsum soils and of seepage waters also accounts for much of the trouble. These matters will be briefly summarized here. In regard to the natural conditions which have had to be contended with, it may be well to state that the settlers were not well prepared to meet these new and untried conditions. In settling the valley the land company attracted a large number of settlers from abroad as well as from various parts of this country. Many of the persons locating in the valley were entirely unacquainted with the exact conditions ijre- vailing in the Pecos Valley, and, indeed, unacquainted with the methods of irrigation in general. The settlers paid the company so much an acre for the use of the water and unfortunately felt that the more they used the more of value they were getting from the company. From these various reasons there was undoubtedly some bad manage- ment which resulted occasionally not only in the financial ruin of the owner, but seriously damaged by seepage waters his own or his neigh- bor's land as a result of over-irrigation. At the present time much of the ownership of the land is vested in the water company and rented to irresponsible Mexican settlers who move from place to place and have not much interest in using the nec- essary means for preventing damage to the land, and who are quite willing to move if crop failures result from their own mismanagement or from any other cause. This social condition must be taken into account in any plan for the improvement and reclamation of the lands. A Mexican settler is not likely to invest money in underdrainage or in any other method of preserving or reclaiming the land and, owing to the uncertain tenure of these tenants, the company naturally hesitates before putting in any expensive improvement of this kind. The con- ditions, therefore, are not as satisfactory as might be wished for the most efl&cient work in the reclamation and general improvement of the lands. Xext to the ownership of the laud and the labor questions, the most important cause of the trouble in the larger portion of the Pecos Valley OPERATIONS IN THE PECOS VALLEY, NEW MEXICO. 19 is the high salt content of the irrigation water, especially in certain seasons. At Eoswell the i)rincipal water supply contains about 76 parts of sol- uble matter in 100,000 parts of water. At Hagerman this is increased to about 200 parts; at Carlsbad to 240 parts; at Florence to 280 parts; Eed Bluff, 31G parts; at Pecos City, Tex., to 400 parts; and below Pecos City to over 500 parts. Five hundred parts of soluble matter in 100,000 parts of water, when added to the soils of the Pecos Valley, may be taken as the extreme limit of endurance for plants, while 250 or 300 parts mark the danger point at which the results of the use of the water are very uncertain. It should be stated that only about 50 per cent of the solid matter contained in the Pecos waters is harmful, and the above figures should be divided by two to obtain the real concentration so far as really harm- ful salts are concerned. The inert soluble matter consists of gypsum and carbonate of lime held in solution through the influence of other more soluble salts. These substances crystallize out on the evapora- tion of the water and are not readily dissolved thereafter. These are, therefore, not liable to accumulate in the soil in a soluble form, and for this reason due allowance should be made for the proportion they make of the total amount of soluble matter in the water. The limit of endurance for most cultivated plants in a water solution is about 1 per cent or 1,000 parts of the readily soluble salts in 100,000 parts of water, but it must be remembered that in field culture the water is applied to soils already containing more or less of these salts, and also that evaporation and consequent concentration immediately set in after the application of water. It was found at Carlsbad that about 300 parts of soluble matter per 100,000 parts of water marked the extreme limit of safety of the use of water at that place. In the use of water for irrigation purposes so close to the limit of plant endurance, as occasionally occurs at Carlsbad, there are certain precautions which can and should be taken to insure the safety of the crops and the permanence of the fertility of the soil. In the arid climate of New Mexico where evaporation is so excessive the actual amount of evaporation becomes an important factor in deter- mining the condition of the water. It is stated that tlje evaporation at Carlsbad amounts to about 10 feet in depth in the course of a year. Evidently, therefore, the deeper the reservoir in proportion to its area the more economical it is as regards the condition of the water. The same bulk of water in a reservoir 20 feet deep with perpendicular sides would lose just half as much by evaporation as if the reservoir were 10 feet deep and the surface area twice as great. There are, of course, engineering conditions which have to be considered in the construction of such reservoirs. As the absolute evaporation from an equal area of water surface is just about the same where the reservoir is full or nearly empty, the •20 FIELD OPERATIONS OF THE DIVISION OF SOILS. water in ;i partially filled reservoir becomes more concentrated as a result of this evaporation than where the reservoir is full. So far as the condition of the water is concerned, therefore, it is advisable to keep the storage reservoir well filled. lu cleaning or repairing the reservoir the water is occasionally drawn off in the fall or winter at the beginning of the dr^' season when work is slack and the danger of floods is at a minimum. If, however, the water has to stand for a long time in this low stage the evaporation has much more effect upon the concentration than if the surface had not been lowered. If the usual rains are delayed and this water has to be used to start the crop, it may be so concentrated as to be unfit or unsafe for use. This is a matter which requires the careful consideration of the engineer in charge. In a large drainage area, such as is found in the upper part of the Pecos Valley, the alkali salts usually come to the surface and form a crust over extensive areas during the dry season. With the first floods much of this salt is carried down into the river and, as a consequence, the first waters of the spring floods at Carlsbad contain a dangerously high salt content. If the reservoirs have been low and the waters are already concentrated the first flood waters may make the waters of the reservoirs unfit foi' use until further rains and floods have come down. Where possible, under conditions of this kind, it would be advisable to divert the first flood waters and not allow them to enter the reservoirs at all. In many places, of course, this would be an engineering feat of so much difficulty and expense that it could not be considered, but where it is possible, under conditions such as prevail at Carlsbad, it would be very desirable. With the irrigation water so near the limit of endurance as some- times prevails in the Pecos Valley it would be desirable to have a record kept and daily or weekly notifications sent to the users of the water of the actual conditions as regards the salt content. With one of the electrical instruments in use by the Division of Soils, the total salt content of the water could be determined in a few moments, at the reservoir, in a flume, in the canal, or in the river before it enters the reservoir. The total salt content thus determined, representing the condition of the water, should be communicated by telephone or other- wise to the principal landowners and users for their information and guidance. It may be perfectly safe to use water of a relatively high salt con- tent on certain well-drained soils when it would be ruinous to allow the same water to be used on a poorly-drained soil containing a high salt content. Furthermore, the previous condition of the soil and the kind and age of the plant has much to do with the safety in applying water which approaches the limit of crop endurance. The soil map accompanying the report of Messrs. Means and Gard- ner shows that there is no e.Kcess of alkali in the Pecos sandy loam, NEW MEXICO. 21 whicli includes most of the cultivated land in the valley, except in a few draws and low places. With this map and the alkali map it should be possible to determine the limit of the salt content of the "water which it would be safe to use on any locality given on the map, as well as for the different plants at their different stages of develop- ment. This is a matter which should receive the thoughtful attention and consideration of all who are interested in the development of agriculture in such a community as this. At several points along the Southern Canal, in the Carlsbad dis- trict, the canal bank is thrown up on the lower side and the water is allowed to rise and flow over an area of from a quarter of an acre to sev- eral acres in extent on the upper side, forming a small lake. This is avowedly done to save the labor of making embankments and for increased storage purposes. It is an exceedingly pernicious thing to allow, however, for two reasons ; namely, it presents a relatively large surface for evaporation and in this way increases the concentration of the solution to a marked extent, and the seepage from the canal is very greatly increased by allowing the water to spread out over such an area. It is very common to find small ponds or lakes some distance from the canal, formed and supplied by the seepage waters from the canal. The ordinary seepage from the Southern Canal probably ac- counts for a large amount of the trouble from seepage waters in the Carlsbad district. This trouble is confined largely to the conglomerate and gypsum areas, but it is unquestionably felt to quite a marked extent in other soils. It was found that in one place in the gypsum area the canal lost about 20 per cent of its volume in a distance of about IJ miles. This is going on all the time the canal is flowing. There is no question but what the excessive subirrigation of the gypsum lands, and probably of the other subirrigated lands shown upon the under- ground water map, is due to the seepage from the canal and its later- als, rather than from overirrigation by the farmers. This matter of the storage ponds along the canal, which has just been referred to, is responsible for a great deal of this unnecessary injury. In Bulletin No. 14 of this Division, on the Alkali Soils of the Yellow- stone Valley, attention was called to the fact that lands receiving the most careful treatment by the people living on them might be ruined by seepage waters and alkali caused by the seepage from canals or by the improper use of water on a neighboring place, perhaps several miles distant. Many instances of this kind have been seen, not only in the Yellowstone Valley, but in the Pecos Valley, which is now under consideration. It seems to me that this is a matter for the States or for Congress to take up with a view that provision be made by statute that reasonable care be taken to protect ditches and canals fi-om undue loss of seepage waters and to provide that reasonable care be exercised in the use of irrigation waters. Provision should be made for the recovery of damages through civil suits in case of injury to 22 FIELD OPERATIONS OP THE DIVISION OF SOILS. property through such sources. It is quite as important for the States to exercise such police power in xirotecting a man's property from destruction by seepage waters from a canal or from injury by the excessive use of irrigation water by a neighbor, as to protect his property from depredations of any other kind. It is of the utmost importance, in a locality similar to the Carlsbad district, where the seepage from the canal is seen to be doing so much injury, that the canal company be required to provide against this loss of water through seepage, and so protect the surrounding country from the dangerous effects of the seepage waters which are seen to occur. Particularly is this necessary where the canal goes through a loose gravelly area or through the gypsum soils, which have the peculiar property of transmitting seepage waters so readily and for such long distances. There are various ways in which such protection can be aflbrded, ways that are voluntarily used by the water companies in districts where water is scarce. It is neither right nor reasonable to permit the wide destruction of property values simjily because water is plentiful and is in itself cheap. Where the cause of the trouble is shown so clearly as in the Carlsbad district there should be no hesita- tion in providing adequate protection to the community. The maps show that with few exceptions the soils of the Carlsbad district are free from excessive quantities of alkali. With the knowl- edge gained through this investigation of the dangers inherent in the relatively large salt content of the waters, to the danger from seepage from the canals and from overirrigation, and the nature of the soil, it should be possible to provide that such careful methods be used that certain lines of agriculture can be successfully carried on under the prevailing conditions. There are other questions which determine to a considerable extent the economic value of crops in this area. For example, the soils of the valley contain a large amount (from 10 to 20 per cent) of carbonate of lime. It is the experience in certain localities that apple trees will not grow well with a subsoil containing so much as 20 per cent of lime. In the matter of sugar beets also, where the water used contains an undue amount of salt, a large proportion of the sugar fails to crystallize, and is lost. There are other qu tions quite as important as these to be con- sidered, but they belong more particularly to some of the other divisions of the Department of Agriculture. One thing should be said in connection with the Carlsbad area, which seems rather anomalous in view of the statements of other investigators, namely, that with a water supply so near the limit of crop endurance as this becomes at times and in those areas in which there is already a large accumulation of salts, that economy in the use of irrigation water, which is generally recomtoended in alkali regions, is one of the worst methods which can be practiced. Where the soil contains a relatively large amount of salt and but little water cou- OPERATIONS IN THE PECOS VALLEY, NEW MEXICO. 23 taining much salt is frequently applied, the ordinary evaporation will increase the salt content of the soil to such an extent that crops can no longer survive; whereas if adequate drainage is provided, and a large amount of water is used, the excess of salt resulting from the evapora- tion of previous applications of water, may be removed, and the soil moisture be maintained at nearly the same concentration as the water supply. It is advisable, therefore, over certain localities, at least in the Carlsbad district, to provide underdrainage where necessary, and then use relatively large applications of water, rather than frequent small applications. If the drainage is adequate, it may be that an occasional flooding in the winter would leave the land in good condition for the coming season, and that during the season frequent small appli- cations can then be used to advantage. A portion of the town of Carlsbad lies directly under one of the con- glomerate banks. The seepage through this from the canal has raised the water surface to within 2 or 3 feet of the surface of the ground. As a consequence of this the shade trees have suffered, and many of them have died. An attempt was made to improve the conditions by stop- ping the irrigation altogether. This made matters worse. The water in the soil concentrated so rapidly and to such an extent that the trees began dying off quite rapidly until, on the advice of Professor Skeats, irrigation was once more resumed, and the conditions were ameliorated. The irrigation in this case should be supplemented by underdrainage, which would not only lower the water level in the soil and secure better aeration, but it would secure better control of the salt content of the soil by occasional flooding. Some observations by Mr. Gardner on these soils are interesting and instructive. A series of tubes was inserted into the soil down to the water level, so that the fluctuations of the water level could be observed with a rod or float. In addition to this, a series of electrodes was put into the soil at intervals 2 or 3 inches apart down to stand- ing water. It was possible by these means to watch the effect of irri- gation on the soil moisture. Before irrigation the level of standing water was 3 feet. Within thirty minutes after the irrigation water had reached the place both the tubes and electrodes showed the level of standing water to have risen to the surface of the ground. The water was held on for some time after the complete saturation had been effected, and after it was stopped it took several days for the natural drainage to restore the normal conditions and for the water level to sink to its former level. Such conditions must be unhealthy for crops, and if the excessive flooding and complete saturation is necessary to remove accumulated salts, artificial drainage should be introduced to quickly remove the excess of water used, as well as to lower the level of standing water in the soil. The maps show that in the Eoswell district extensive areas of land are already ruined by seepage waters and by alkali. This has unones tionably come from the seepage from the canals and from overirriga- tiou. The area is small and seems adapted to intensive farming, and 24 FIELD OPERATIONS OF THE DIVISION OP SOILS. there is no reason why at comparatively small expense these lands should not be underdrained and the present trouble entirely reuioved. The water supply is so good that with adequate drainage provided, there is no question about the ease with which the damaged land could be reclaimed. The conditions are so clearly presented in the report and accompany- ing maps and the remedies are so obvious that further comment in this place is unnecessary. The investigations certainly point to a most encouraging prospect for success in the Roswell district. Pecos City and Barstow are approximately 80 miles south of Carlsbad. The Pecos Eiver, supplying these places with irrigation water, gathers the uuused, waste, and seepage waters from the Roswell, Hagerman, and Carlsbad irrigated areas, and the salt content of the water is so great as to be beyond the limit of endurance for most plants during the greater part of the time. Nothing can be said, no advice can be given for the successful irriga- tion of lands where the water supi)ly has so large a salt content as at these places. It is just possible that at certain times of the year and in certain states of the water the water supply may be pure enough for successful irrigation, and that by the construction of large storage reservoirs and the diversion of the water when the salt content reaches a certain maximum a small irrigated district could be maintained; but these are questions which would require a detailed investigation extending over a number of years. It would be advisable to have records made of the conditions of the water from day to day for at least two or three years, and not rely upon the single determinations which have been made, although these are supported by the experience of the past years in the practical irrigation of the lands. Mention has already been made of the occurrence of the root-rot disease in the Carlsbad district. So far as can be determined with a single season's observations and without the necessary investigations of the vegetable pathologist, it would seem that the character of the soil with the high content of carbonate of lime, which is known to be prejudicial to certain crops — especially apple trees — and with the water supply so charged with salts as to be near the limit of endurance for crops, the vitality of the crops would be so low that they would become peculiarly subject to diseases which in a more vigorous condition they could readily resist. This is a matter, however, for the Division of Yegetable Physiology and Pathology, and I consider that these records and maps would form a very valuable basis from which they could work out economic problems in vegetable physiology and pathology which others have tried for years to study in the confined limits and under the artificial conditions of pot culture. It is a line of investiga- tion which is well worthy of being followed out, as it is of the utmost economic importance not only in this, but in other irrigated districts of the West. SALT LAKE A'ALLEY. 25 SALT LAKE VALLEY. The Salt LakeTalley is tlie oldest of the moderu irrigation districts and has been noted for the high state of cultivation of the lands by the Mormons. During recent years, however, complaints have been made of the damage by alkali and seepage waters of the lower levels, which were formerly the most productive soils of the valley. The tendency has been for the settlers to move farther and farther back onto the higher benches. The work of the field party was confined to that portion of the val- ley west of the Jordan Eiver and extending to the Great Salt Lake and to the Oquirrha Mountains. This area comprises approximately 250 square miles, or about 160,000 acres. Of this large area only 40 square miles,. or approximately 25,000 acres, are at present under irri- gation, and 50 square miles, or 32,000 acres, are all that have ever been under successful irrigation. About 90 square miles, or approximately 58,000 acres, are above the canals and are at present not available for irrigation. There are about 125 square miles, or approximately 80,000 acres, in a vast level stretch of country west and north of Salt Lake City, extending to the Great Salt Lake. Of this it has been estimated that about 90 square miles, or approximately 58,000 acres, are capable of improvement through underdrainage or irrigation. The remaining 35 square miles of this area are in flats of heavy clay soil of low eleva- tion, filled with alkali, which it would be inadvisable or impossible to underdrain. The water supply of the Jordan Eiver is reasonably pure and is well suited for irrigation purposes. The most recent canals were taken out near the Jordan Narrows and have nearly the same composition as the Utah Lake. Farther down the river gathers some alkali from seepage waters, but it also gains a considerable quantity of fresh water from mountain streams, so that there is considerable variation in the salt content in different parts of its course. There seems to be an ample water supply to irrigate all the land which could be brought under the ditch, but on account of the high salt content of the soil irrigation has never been successfully practiced on the 125 square miles of nearly level land west and north of Salt Lake Gity. Of the 50 square miles which have been under the ditch and suc- cessfully cultivated, about 10 square miles, or one-fifth of the whole area, have been ruined by seepage and alkali salts from the higher levels. This abandoned land is now wet and swampy, and about 1,000 acres are actually covered by large lakes which have been formed by the seepage from canals and irrigated lands. These abandoned lands were originally the most fertile portion of the area, and being for the most part nearer the city had a value probably of not far from $80 to $100 per acre. This damage is gradually extending back toward the bench lands. It could readily be stopped with more care in the con- struction of the canals and in the use of water supplemented by thorough drainage. 26 FIELD OPERATIONS OP THE DIVISION OF SOILS. There is a scheme ou foot at the present time to coustruct a high- level canal, taken out from the Jordan Narrows, to be supplied by hydraulic pumps, which would bring under the ditch a large portion of the 90 square miles :it present above the irrigating canal. This would increase the danger, however, to the lauds at present under irrigation, and steps should be taken to protect these lands and prevent the encroachment of the seei)age waters and alkali on the lower levels. This is a matter which could easily be accomplished at a moderate cost. It would necessitate more careful work in the construction of the canals, to prevent seepage, and adequate drainage on the lower levels. The greatest problem, however, in the valley at the present time is the reclamation of about 100 square miles of the land lying adjacent to and west of Salt Lake City. There seems to be no question about the feasibility of the engineering problem in underdrainage there, or about the efficiency of underdrainage iu the reclamation of these lands with the water supply so good and so abundant as it is. The cost of under- drainage would be no greater than for similar lands in the States of New York, Ohio, or Illinois; while the value of these lands, being so near Salt Lake City, would certainly be as great as in those States where underdrainage has been so extensively carried on and is con- sidered so essential in the cultivation of the land and the maintenance of its fertility. There would thus seem to be a fine opening as a com- mercial enterprise in the development of these lands through under- drainage. There is a mean average fall from the ridge running in a northwesterly direction from Salt Lake City of about 2 feet to the mile down to the lake. This would be ample for the main drainage canals, as the irrigation canals are frequently built with about half this fall per mile. The drainage of these lands would be rendered comparatively easy through the numerous draws already existing, which are from 6 to 8 feet below the general level of the surface into which the drains could empty. Good tile clay is abundant in the vicinity and tile could be made and delivered to the farmer at a reasonable cost. Over such an extensive area as this the drainage systems should be well devised and should be a community affair rather than constructed through individual efforts. The cost of underdraining these lands with tile is estimated at from $ 10 to $20 per acre, depending upon the texture of the soil and the distance apart of the drains. These lands at present have no value or only a mere nominal value for grazing purposes. The irrigated lands of the locality are worth from $60 to $80 per acre. If these lands adjacent to Salt Lake City were reclaimed through under- drainage they would be worth certainly from $60 to $80 per acre, and much more than this if held as suburban property. This reclamation work should be carried on as a community enterprise, aided perhaps by county or State credit, as has been done in many sections of the country where similar enterprises are undertaken. THE CONNECTICUT VALLEY. 27 These are tlie great problems presented in tlie Salt Lake Valley, problems presented very clearly in the report prepared by Messrs. Gardner and Stewart. Prom the evidence there presented, and espe- cially from the maps, the actual conditions in the valley may be clearly understood. The soil map shows the distribution of the different types of soil, the Jordan sandy loatn forming by far the greatest extent of valuable land. This soil is not originally very alkaline and it is easily drained and improved. The Jordan loam is rather more difficult to deal with, while the Jordan clays, from their sliglit elevation and the impervious nature of their material and the large accumulation of salts in them, would present great difficulties and would perhaps prove impossible of reclamation. Where this Jordan clay occurs at a reasonable distance from the surface, however, it undoubtedly adds strength to the lands. Where it comes too near the surface it is an element of danger, as it impedes the drainage of the soil. The relation of these soils to drainage, seep- age waters, and alkali should be very carefully studied by the farmer. The alkali maps and the underground water map show very clearly the conditions prevailing over the whole district. Attention should be particularly directed to the very large accumulation of alkali in some of the areas north of the Twelfth street road. Also the remarkable accumulation of sodium carbonate or black alkali should be considered. This sodium carbonate occurs in places to as great an extent as 3 per cent in the surface foot of soil, while the composition of the crust which forms over the surface indicates at least a possibility that it may be sufficiently pure for commercial purposes. One interesting thing in connection with this area is that calcium chlorid has been found to occur in very large proportions in some of these crusts, and strontium is also present in appreciable amounts. No attempt was made to investigate the economic importance of these salts, but the indicatioTis are that it would be worth while to look into the matter further from an economic point of view. The investigation of the hardpan forming on the shores of Great Salt Lake is interestijig and valuable. It seems to me that the whole matter of the occmrrence and formation of hardpan in soils is very nearly solved, and that before very long this troublesome question in agricultural practice will be fully understood. The observations of Messrs. Gardner and Stewart and the paper by Dr. Cameron have contributed important facts and suggestions toward the solution of this problem. THE OONNJiO'riCTJT VALLEY. The report of Mr. Dorsey on the soil survey of the Connecticut Valley, with the accompanying map, presents so clearly the conditions prevailing there that little comment is necessary at this place. Further- more, it is but the basis of a very extensive and comprehensive 28 FIELD OPERATIONS OF THE DIVISION 01' SOILS. investigation by this division, authorized by Congress in the present appropriation bill. The map shows the distribution of the soils of the valley, which are described with all necessary detail in the accompanying report. The sides of the valley are formed for the most part from the glacial deposits of Triassic sandstone, and in the northern part of diabase. The soils of the valley proper are sediments which have been washed over and assorted in the great lake which is supposed to have covered this area in prehistoric times. Some of the soils occur in well-deiined terraces, which formed the shores ()f the old lake, or which were formed subse- quently by the river and streams. Over much of the area, however, these terraces are ill-deflued or entirely lacking, and, from the differ- ences in elevation of the same soil formation in different i)arts of the valley, there are even evidences to disprove the terraci- theory of the physiography of the country, Certain it is, however, that the soils were laid down by water, and that in so doing they were sorted out in various grades of fineness. Beginning with the present meadows, which are composed of very fine sand and silt, the Podunk region is in a well-defined terrace elevated about 20 feet above the meadows and is composed of one grade coarser material, but still so fine as to be just distinguishable by the eye. The Hartford loam, forming the principal tobacco soil, in extent at any rate, is a grade coarser than this, while the Windsor loam, believed to be the original bottom of the old lake in its shallowest portion, is verycoarsesand.containingsomegravel. These Windsor sands produce the finest wrapper leaf when the season is favorable, but a good crop is secured only one or two years out of five. As Mr. Dorsey points out in his report, the one great trouble with the Connecticut tobacco is that it does not conform to the present require- ments of the cigar trade. The leaves are too large, the veins are too large, the base of the leaf is too glossy and lacks texture and style, while the color of the leaf is far from uniform. An attempt is soon to be made to secure a radical change in the type of the leaf by close planting, allowing many more leaves to the stalk, by very rapid growth, by shading, and possibly by irrigation. These experiments with the Connecticut tobacco will be undertaken in the hope of producing a leaf approaching more nearly the Sumatra type of wrappers, this type being generally accepted in this country as the standard for cigar wrappers. With the intensive cultivation that this will require, it is quite pos- sible that these Windsor sands may be looked to for the finest wrapper leaf. I am of the opinion that even with the present style of leaf it would pay to irrigate these lands where this coul I be done easily and cheaply, in order to secure a crop four times out of five at least where now it is only possible to obtain one or two crops out of five. The Hartford loam is decidedly a safer soil and can be relied upon to produce a fairly good crop of the Havana seed leaf variety each year. CECIL COUNTY, MD. — FIELD METHODS FOR A SOIL SURVEY. 29 This tobacco has a peculiar " seedy" taste which it is desired to get rid of and it is possible that this may be accomplished through selection and breeding, if not by radical departures from the present methods of cultivation and fermentation. Such questions as these will form the basis of an extensive line of investigations already outlined, requiring several years of systematic work. CECIL COUNTY, MD. The soil work for the Cecil County sheet has been completed during the present season, but the topographic base map has not yet been engraved and will not be available for publication before the fall of 1900. Photographs of the pencil copies were kindly furnished by the United States Geological Survey for the actual field work and transfers will be obtained as soon as the plates are engraved. The soil map will be published in cooperation with the Maryland geological survey. LABORATORY WORK. The paper by Dr. P. K. Cameron, soil cliemistof the division, cooper- ating with the Division of Chemistry, together with the one by Mr. Lyman J. Briggs, physicist of this division, shows some important advancements made in the methods of soil investigations, which have very materially increased the efficiency of our field parties and have made them quite independent now of the laboratories, except so far as the laboratories will be used to investigate and explain the basis upon which the soils have been classified in the field in order to show as fully as possible the physical and chemical peculiarities of the ditt'erent soils. The electrical methods of determining the moisture and salt content of soils are based upon the conception of the soil moisture as being a solution of a mixture of more or less difficultly soluble substances. Dr. Cameron has carried this further in the application of the modern theories of solution to the study of some of the most difficult soil problems. As a result of this, new methods of field analysis of mineral substances have been devised, and the occurrence of certain mineral substances in the alkali soils and crusts has been explained with sug- gestions for the more rational treatment of the land, and, lastly, the occurrence and mode of formation of hardpan has been explained more clearly and more rationally than ever before. The work naturally leads up now to a study of the interesting and important subject of absorption by soils and the chemical and physical changes induced by fertilizers. FIELD METHODS FOR A SOIL SURVEY. After years of careful work the methods adapted to the field survey of soils have been perfected, so that now the field parties are quite independent of the laboratories, at least in the preparation of their 30 FIELD OPERATIONS OF THE DIVISION OF SOILS. maps. A brief descriprioTi of the field methods in use by the division at this time may be of value to those who are interested in tlie work. In the tield classitication of soils all features are taken into consid- eration which appear in any way to influence the relation of soils to crops. The classification is based mainly upon the physical properties and condition of the soil as determined by the soil expert, but it is not based solely on this. Any chemical feature, such as deposits of marl, of highly calcareous soils, or of highly colored soils, is considered, as well as the character of the native vegetation and the condition of the crops. The topography of the country is often a very safe guide in outlining the boundaries of soil conditions. For this reason it is very advisable that there should be reliable maps of the different districts to base the soil work on and that these maps should show the important topographic features. In order to make the maps of the greatest possible local value it has been decided to publish the soil maps, so far as possible, on a scale of 1 inch to the mile. Local variations in the character of the soil of less than one-fourth of a mile in extent are generally ignored, unless this variation constitutes a very prominent feature, such as a strip of meadow land along a stream, or unless there are a number of small areas by which a certain character is given to the district. For example, if rocky areas occur, small in each case, but extending over large areas, they should be indicated in some way upon the soil map. The basis for the field classification of the soils of the Connecticut Valley may be seen from Mr. Dorsey's report. The meadow land is not only set off as a distinct physiographic feature, depending upon the topography of the country, but it is a very fine sediment of silt and very fine sand, which is easily recognized and is distinct from any of the other soils of the valley. The judgment of the soil expert in the field in a matter of this kind in deciding on the texture of these soils is very reliable. The Podunk loam is a grade coarser than the meadows, and this also is very apparent to the observer in the field. The difference in texture in these soils can be recognized quite clearly in dried samples in the laboratory. The difference in the character of the vegetation and in the relation of crops as seen in the field is very marked. Extending back from the river on either side and at higher eleva- tions is the Hartford loam, which is still coarser in texture, while the Windsor sands are very coarse angular quartz grains with a little fine gravel. The Triassic stony loam has a peculiar Indian red color and contains fragments of stones grading up in size to boulders. The Enfield loam is apparently a deposit of the Hartford loam immediately upon the Triassic stony loam, with an average depth of about 18 inches. The Chicopee gravelly loam, covering a large area in the northern part of the district, has so much gravel as to form a very conspicuous feature. These features, upon which the classification of these Connecticut soils FIELD METHODS FOR A SOIL SITKVEY. 31 is based, are quite plain and distinct enough in the field to outline the soil areas. In the preliminary work in Florida, which was mentioned in Bulletin No. 13 of this division, it is evident that the classification of the soils there had to be made mainly from the distribution of the native vegeta- tion. The diffetent classes of soils there varied but little in their physical and chemical properties, but there is a very great difference in the native vegetation and in the adaptation to crops. The reason for this is not at present understood, but it is mentioned here to show that the basis for the classification of soils differs under different cir- cumstances. The survey of the alkali soils of the far West is much more difficult, as it involves other determinations and observations in the field. There should always be a soil map, an alkali map, an underground water map, and where sodium carbonate exists in appreciable quantities a separate map should be made showing the distribution of this pernici- ous substance. It is very important that the data for the soil map be first collected and the soil districts outlined on the best available base map. This will enable a much more intelligent stndy of the alkali problem to be made than if the distribution of the soils had not pre- viously been determined. An example of this (an be seen in the Salt Lake sheet which accompanies Mr. Gardner's report. As a rule the clay soils on the flats and draws of the great alkali plain west of Salt Lake City contain the greater amount of alkali, the loam soils next, then the sandy loam, and finally the sandy soils. This is partly due to the texture of the soil and the influence of this texture on the drainage, and partly to the physiography of the country as determining the drainage question. Having located these soils and considered the influence of the topog- raphy, a comprehensive study of the alkali conditions is much simpler. In the preparation of the soil maps only such conditions as are ap- parent in the field, such as the texture as determined by the feel and appearance, the depth of soil and subsoil, the amount of gravel, the condition as to drainage, and the native vegetation or known relation to crops, are mapped. Having determined upon the final classification of the soils of a locality, each well-defined area is established as a class and given a local name. No attempt will be made for the present to correlate a loam soil in the Connecticut Valley with a loam in the Susquehanna Valley in Pennsylvania unless the two are very clearly identical in origin, in character, in relation to crops, and under essentially the same climatic condition. If there is any apparent difference it is preferable to give each a local name and describe each separately in the most careful and exhaustive manner. Having decided upon the classification, samples should be taken from each soil formation in order to study the texture and chemical 32 FIELD OPERATIONS OF THE DIVISION OP SOILS. composition. For this purpose each area should be represented by at least eight or ten samples so selected that they will represent the aver- age conditions of the area. The depth to which these samples should be taken depends somewhat upon the locality. lu the eastern district the samples are usually taken to a depth of 3 feet and occasionally deeper. In the western district, where the soils are more uniform and where the alkali conditions are important, the samples are usually taken to a depth of 6 feet and occasionally deeper. The samples representing any particular formation should be fully described and such a number of them taken for examination, both physical and chemical, as may be thought necessary to bring out the striking features of the soil. The information so acquired will be used, not as heretofore to base the classification upon, but only to describe the samples mentioned in the text and to show the relation of the different soils of the area. In the construction of the alkali maps it was formerly the practice in this division to take borings at more or less regular intervals along section lines and afterwards construct a map from the field notes after they had been corrected and the figures leduced in the laboratory. Much time can be saved, however, and greater accuracy secured by pre- paring the maps in the field as the work progresses, as is always done in the case of the soil maps. Having outlined the soil areas, each area should be studied in sufS- cient detail to determine its condition as regards alkali. Very fre- quently, from the character of the soil and the topography of the Luid, the conditions of alkali over extensive areas can be very satisfactorily determined with a very few salt determinations. The alkali conditions of each soil should therefore be studied in this way. It is usual to make borings for alkali work to a depth of C feet or more. It has been found, however, that the data so collected is really more than is necessary for the construction of an alkali map. It is now considered sufficient to examine the first, third, and fifth foot of the boring for alkali salts and neglect the second and fourth foot. In plotting the results on the map it has been customary to locate the borings on the base map and faintly indicate the salt content of the first, third, and fifth foot. Having determined the limits which it is desired to show upon the map, lines are drawn around the figures in such a way as to separate the areas falling within or without the limit- ing values. It has been found better to use judgment in the placing of these lines, based upon the salt content of these three depths, rather than to accurately outline the conditions of the surface foot or of any given depth below the surface. Very frequently from some local cause, such as a slight elevation or a ridge formed in the cultivation of the soil, etc., the surface foot may contain a small amount of alkali, while the lower depths will show an excessive amount. These lower depths thus serve to check any error that may occur from a consideration of the surface conditions. Furthermore, even if the surface foot is shown FIELD METHODS FOR A SOIL SURVEY. 33 to be free from alkali over a considerable area which has never been irrigated and excessive quantities are shown in the other depths, it would be entirely misleading to show the soil free from alkali as a sur- face map would, as trouble would very soon be apparent after irriga- tion was started. It is not considered advisable to publish maps show- ing the alkali conditions at different depths, but far better in arranging the map to rely upon the judgment of the soil expert in outlining what may be considered as safe, dangerous, or worthless soils, so far as alkali salts are concerned. This work has formerly been done in the office after the salt content corresponding with the actual resistances in the field had been reduced. The present practice, however, is to map these salt areas in the field in the same manner as the soils are mapped. A salt map is usually shown in colors, these colors indicating the soils having a very small and safe alkali content, those in which the alkali is too strong for crops, and intermediate grades, which may be considered dangerous and which would have to be handled with care and judgment. In the work so far this medium grade has a limiting value of one-half of the maximum — that is, if the maximum amount for a crop of any particular kind of alkali is 0.50 per cent, the lower limit of the danger line would be placed at 0.25. This is an arbitrary figure, but one which seems to be justified by the facts. In testing the limiting value it is customary to look for a crop such as alfalfa or wheat, which is showing the effect of alkali in some portion of the field. Very frequently the middle of the field may begin to die out while the edges are still flourishing, or perhaps one corner will show the effects of alkali before the rest of the field is affected. By careful work on such a field as this the limiting value can be determined. The electrical resistance of the eoil at various points in such a vicinity should be determined, corrected for temperature, and the limiting value be selected from the resistance figures so obtained. As a result of the work this season, the following directions have been issued for the guidance of the field parties : DIRECTIONS FOR MAPPING ALKALI SOILS IN THE FIELD. The contour intervals for tho alkali maps are to represent, respectively, 0.20, 0.40, 0.60, 1, and 3 per cent of salt in the dry soil. The maps are to be constructed in the field directly from the resistances. The work is to be standardized in each district in the following way : Take eight or ten crusts, including the top inch of soil; or, if crusts can not be obtained, take the strongest alkali soils from different places over the whole area. Fill a large cnp or tumbler about one-third full with a crust or soil, using more or less according to the richness of tho material, and nearly fill the tumbler with distilled water. Stir vigorously and allow it to subside for a short time. Treat the eight or ten crusts or soils in the same way. Determine the electrical resistance of the solutions in the cell. Take an amount of the strongest solution eriuivalent to at least 200 cc. having a resistance of 5 ohms, and add to it a volume of each of the others proportional to the resistances determined. This mixture, containing approximately equal quantities of salt from the eight or ten localities selected, is H. Doc. 399 3 34 FIELD OPERATIONS OF THE DIVISION OF SOILS. evaporated to dryness on a common range in a graniteware saucepan. Before the salts begin to crystallize out and when the clay and organic matter are well floccu- lated, filter and evaporate the clear filtrate, stirring at the last to prevent caking. Gently heat the residue to drive off the water of crystallization of the sodium sul- phate and sodium carbonate. If the original solution can not be filtered clear at any time before the salts begin to crystallize out, a second evaporation may be necessary. A 10 per cent solution should be made of this salt in distilled water. The elec- trical resistance of this 30 per cent solution in any cell, divided by 0.24, will equal the resistance of a saturated sand or sandy loam soil in the same cell when com- pletely saturated and at a temperature of 60° F., when the soil contains 3 per cent of salt. This solution is then to be diluted and the resistance determined at various concentrations, corresponding to the limiting values of the soil map for four grades of soil. The table for temperature reduction have been published in Bulletin No. 8 of the Division of Soils. The dilutions are as follows, the figures representing cubic centimeters of the 10 per cent solution to be diluted and made up to 100 cubic centimeters: To obtain limiting values. Volume of 10 per cent Bolution in 100 cc. Bait in soil. s. and S8C. 8C. sec. c. and he. Per cent. ec. cc. cc. cc. 3.00 100.0 79.5 71.4 66.0 :.oo 33.3 26.6 23.8 22.2 .60 20.0 15.9 14.3 13.3 .40 12.0 10.6 9.5 S.9 .30 6.0 5.3 4.8 4.4 Kesistances to be re- duced lo 00° f. and divided by . 24 .275 .29 .30 The result will give the cell resistance at 60° F. corresponding to the limiting values, to be inserted in a suitable table. To correct for any lime sulphate which does not redissolve, take the resistance of the solution before evaporating. Then evaporate in a separate dish a measured portion, such as 100 cc, heat the residue to char the organic matter, dissolve, make up to the same volume as before, and take the resistance. Any difference in resist- ance will be considered due to salts having gone out of solution, and all resistances used in the limiting values are to be decreased by the proportional increase in resist- ance so found before being entered in the table. Or the weight of salt for the 10 per cent solution may be increased in proportion to the differences in resistance. DIRECTIONS FOR ESTIMATING SODIUM CARBONATE AND CHLORIDS IN SOILS. Take a known volume (or weight) of saturated soil, wash into a 250 cc. flask, and fill to the mark with distilled water. Take 50 cc. of the solution (a slight turbidity will not matter) and titrate with N/10 acid potassium sulphate, using phenolphtha- lein as an indicator. Then add a few drops of potassium chromate as an indicator to the same solution and titrate with N/10 silver nitrate. One CO. N/10 Na2C03 = .005266 grams NajCO.,. One cc. N/lO NaCl = .005806 grams NaCl. Construct the sodium carbonate map in the field from the volume of solution used. Limiting values will be 0.3,0.2,0.1, and 0.05 per cent of dry soil. The limiting values FIELD METHODS FOE A SOIL SURVEY. 35 for each dish are found, in the following way : Multiply the volume of saturated soil, represented by the solution taken for titration, by the numbers in the following table : NajCOsineoU. s. and 880. 80. sec. c. and ho. Per cent. .30 .832 .752 .720 .689 .20 .454 .602 .480 .459 .10 .277 .251 .240 .230 .05 .138 .125 .120 .115 The results so obtained are the cubic centimeters of N/10 solution of sodium car- bonate corresponding to the limiting values, to be inserted in a suitable table. If it is desired to reduce the volume of N/10 AgNOs to per cent of NaCl in dry soil, the following formula may be used: V X .005806 V'K V = CO. N/10 AgNOs solution used ; V = volume saturated soil represented in amount of solution titrated; K = constant for type of soil, as follows : 8. and ssc. (sand and sandy loam) ^ 1. i6 ; sc. (loam) ^1.32; sec. (clay loam)=1.26; u. and he. (clay and heavy clay) ^1.21. ' If the resistances are close to the limits, a temperature reduction should be made in the field and a decision reached there as to which class the soil should, be assigned. This greatly simplifies the work, lessens the calculations, and insures greater accuracy, because with the soil areas in view and with the general topography of the country the boundaries of these limiting values can be more accurately drawn than if the work were subsequently done in the laboratory from the field notes. In like manner, the sodium carbonate map caoi be constructed in the field from the amount of standard solution of acid potassium sulphate required to neutralize the alkali in a given volume of saturated soil. In the work so far undertaken Hilgard's value of 0.1 of! per cent of sodium carbonate has been taken as the limiting value for crops. The minimum value for the danger line is taken arbitrarily as 0.05 of 1 per cent. The amount of acid potassium sulphate required to neutralize a soil containing 0.1 of 1 per cent of sodium carbonate can easily be determined for any given volume of saturated soil and the survey can then proceed upon this basis in the classification of the black alkali conditions. No calculations are therefore required, except the ordinary standardization of the solutions and the determining of the volume of the measuring dish and the amount of dry soil which it contains. When these are once accurately determined, the work can proceed quite rapidly. The construction of the water map is, of course, the simplest of all. Data are secured of the depths of all surface wells and wherever water is encountered in boring for samples. From the data thus secured the areas are traced upon the base map. A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. By THOS. H. MEANS AND FRANK D. GARDNER. INTRODUCTION. The southeastern part of the Territory of New Mexico comprises three distinct topographic features. The mountains of central New Mexico, including the Guadalupes, White, and Sacramento mountains, with the Santa Pe Range, mark the western boundary of the great series of plains extending westward from the mountains of Arkansas. On the eastern boundary of the Territory of New Mexico the Great Staked Plains are the most important feature. These plains form an elevated plateau with a uniform slope to the east. The western escarp- ment of the plateau forms the eastern limit of the Pecos Valley. The Staked Plains once, without doubt, extended to the foot of the moun- tains, with their western boundary somewhere near the present foot- hills which bound the east slope of the mountains, but the gradual uplifting of the western edge of the plain, together with excessive ero- sion along the base of the mountain, formed what is now known as the Pecos Valley. The present topography of the land is the result of the erosion and sedimentation of the stream in this valley — the Pecos River. At some period in its history, a time perhaps corresponding with the glacial period of the North, the Pecos carried much more water thau at present, and during this time dams formed along the river either of hard ledges of rock or, by the filling up of narrow gorges with drifting material, caused the water to back up into shallow basins extending over great areas of country. In these basins or inland lakes the waters descending from the mountains deposited large quan- tities of sediment. The most pronounced basin of this type has been recognized and named the Tayah Basin. This basin lies on the lower Pecos, with its northern extremity near the Texas-New Mexico line, and extends for an undetermined distance to the southward. Between the Delaware and the Black rivers, in a stretch of broken country, a second dam probably existed, and this obstruction backed the water. up to some distance beyond Carlsbad. The sediments enter- ing this basin were largely from calcareous rocks of the Guadalupe Mountain, and the soil formed from these sediments, weathering under the arid conditions of New Mexico, carries large quantities of carbon- ate of lime. 36 INTRODUCTION. 37 Between Carlsbad aud Seven Eivers the Pecos flows through rough country along the foothills of the Guadalupe Mountains, aud in this section of the river another obstruction once existed. The lake formed by this dam extended as far north as Roswell and an undetermined distance westward from the Pecos. In this basin were deposited the sediments which form the soils of the Eoswell and Hagerman farming district. There are many minor basins and features of the Pecos River which were not studied, since their bearing upon the subject in hand is only of secondary interest. The rocks out of which the valley was cut vary in age from the car- boniferous of the Guadalupes through the Permian Eed beds of the upper Pecos, above Eoswell, to the Jura-Trias and cretaceous sedi- ments of the Staked Plains. The carboniferous rocks are composed almost entirely of magnesian limestone, with beds of shaly limestone and thin sandstone. Such rocks form poor soils under the arid conditions existing, since they contain large quantities of lime without much potash or phosphoric acid. The Eed beds consist of red sands and shales, with heavy beds of massive gypsum. Gypsum crystals are common throughout the forma- tion, and indicate formation from inclosed basins of sea water. The Pecos Kiver drains nearly the whole of the southeastern third of the Territory of New Mexico. Eising on the east side of the Santa Fe Eange, the stream flows as a typical mountain stream through the rocks of the mountains; then entering the horizontal rocks of the mesa country the stream assumes a meandering course broken at intervals by gorges and canyons. The general character of the Pecos below Eos- well is a series of basins tilled with lake sediments and separated by rough country and hard rocks, through which the Pecos Eiver is at present cutting. The main tributaries of the Pecos all come from the western side, and they, too, are mountain streams, rising in the White, Sacramento, and Guadalupe mountains. The upper branches of these streams flow throughout the year, but as soon as the level mesa country is reached most of the streams sink into their beds. During times of high water the streams flow throughout their entire courses. The water which sinks along the upper stream courses follows under ground the general course of the rivers and appears along the basins near the Pecos in the form of springs. During its course through the ground the water dis- solves small quantities of soluble matter and most of the springs contain the common alkali salts. Above Eoswell the main Pecos has few tributaries of any size. Prom Eden south small quantities of water flow in its channel throughout the year, though as far south as Eoswell the flow sometimes is hardly more than 50 cubic feet per second. 38 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. At Roswell there are several permaueut sources of supply which flow an estimated quantity of about 200 cubic feet per second. The Beren- dos, a series of large springs, rise from the edge of the large gypsum plains which extend for a distance up the Pecos Valley. The water from these springs has its origin in the crevices and underground channels of these gypsum plains, which form part of the Red Beds, and no doubt comes from the upper Pecos, the streams entering from the west across this gypsum plain. All of the waters coming from gypsum areas contain more soluble matter than do waters from the limestone strata of the underground river basins. This can be accounted for upon the assumption that the gypsum had its origin in inclosed basins of sea water, which always contains calcium sulphate in solution, concen- trating through evaporation. Owing to the small solubility of the gypsum, this is the first salt crystallized out, and even though the water does not cou.. Coarse sand . . . Medium sand.. rino sand Very iine sand. Silt Fine silt Clay Per cent. Per cent. Per cent. Per cent. 1.11 2.55 14.12 38.57 13.17 9.63 13.62 Trace. 3.95 16.45 29 15.18 9.76 17 Trace. 1.52 9.15 34.10 17.34 9.70 18.45 Trace. 1.22 6.97 29.61 19.30 12.10 21.90 Saltsa Loss at 110° C... Loss on ignition. 1.60 2.70 2.75 1.85 5.28 3.85 1.46 4.98 3.94 Per cent. Trace. 2.05 9.76 33.11 19.02 10.35 13.77 1.46 4.59 4.55 aDisflolved in 1} liters of water nsed in mechanical analyses, mainly gypsum. The mechanical analysis of this soil, when first made by the customary method in this division, showed as much as 30 per cent clay. This dif- fered so much from the field observations, for in the field these soils were classed throughout as sandy loams, that the source of the differ- 46 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. ence was investigated. A microscopical examination of the soils showed them to be composed of small conglomerations of clayey mat- ter cemented by carbonate of lime. By agitation with large quanti- ties of water this carbonate of lime was partially dissolved, the clay loosened, and the resulting analysis would make the soil appear heavier than the field judgment would warrant; also it was found that there was a quantity of matter in the soil which dissolved in the water used in the analysis. This was not in all cases alkali, but was more likely to be gypsum and carbonate of lime, both of which are slightly soluble in pure water. This soluble matter is always included in the clay by the method employed in the analysis, and thus served to make the soil appear heavier than it really was. This soluble matter was determined in all cases and subtracted from the clay. The amount of soluble mat- ter is given, but must not be confused with the alkali determination given in other tables. The action of water upon this soil in the field produces nearly the same results as were observed in the laboratory, and where irrigated a difference can be seen in the apparent clay content of the soil. After irrigation continued for some time the sandy loam changes in appear- ance and character to a loam or even to a clay loam. THE ROSWELL SANDY LOAM. This soil differs from the Pecos sandy loam in being a little heavier, though the top foot of soil may be nearly of the same texture. It is underlaid at from 1 to 2 feet by a loam, and this in turn is underlaid at 5 feet by a clay loam or clay. The native vegetation of this soil was originally the same as on the Pecos sandy loam, but since irrigation water has been applied all waste tracts are now covered with salt grass and alkali weeds. The texture of this soil in the field was considered different from the soils of the prairies. The difference was so slight that the results of the mechanical analysis fail to show it, as appears from the following table; still the mechanical analysis is not considered perfectly reliable in these soils for the reasons stated above. Mechanical analyses of soils o / Boswell sandy loam. Diameter. Conventional names. 4099. Eoswell, 3 miles SB. 4108. Eoswell, 5 miles E. 4112. Eoswell, 5 miles SB. Millimeters. 2 to 1 1 to 0. 5 0.5 to .25 .25 to .1 .1 to .05 .05 to .01 .01 to .005 . OOS to . 0001 Salts dissolved mechanical an Loss at 110° C. Gravel Per cent. Per cent. Per cent. 6.35 1.90 15.87 30 14.83 8.90 15.66 Trace. 2.20 17.80 34.22 14.46 8.03 17.10 .40 12.40 32.80 15.65 10.22 15 Pine sand Very fine sand Silt Fine silt n IJ liters of water used in alysis, mainly gypsum L74 3.13 7.15 4.50 3.40 6.25 1.60 2.61 3.60 Loss on ignition ROSWELL DISTRICT — ROSWELL SANDY LOAM. 47 There is little question but that the arrangement of the soil grains plays a great part in the field judgment of the texture. The applica- tion of the irrigation water separates the iioccules and to a small extent gives the soil a much heavier appearance than the mechanical analysis shows. This fact is plainly to be seen in the Pecos Valley soils The Roswell sandy loam may be considered the same as the Pecos sandy loam, with the upper part of the stratum removed, bringing the loam and clay nearer the surface. Since the alkali is more abundant in the lower layers of the Pecos sandy loam, it is evident that the salt should be more abundant in the Eoswell sandy loam, which represents the lower strata of the Pecos sandy loam. Such was, no doubt, the case, though it has been found diflScult to secure a sample of this soil which has not been irrigated. The following table represents the soil as near its original condition as it is possible to find : Salt content of Boswell sandy loam iefore irrigation Depth. 67. Depth. 67. Feet. Per cent Feet. Per cent. 1 0.07 5 0.36 2 .27 6 .25 3 .41 7 .20 4 .54 8 .20 After irrigation water has been applied to this soil a very different state of affairs is found to exist. The greater part of the soluble matter is dissolved in the irrigation water, and should this water be able to sink into the lower subsoil and flow off into the country drainage, most of this salt would be entirely removed from the soil and do no further damage. This soil, however, becomes heavier and more compact in its lower strata, and consequently the percolating water finds difficulty in penetrating to the stream channels. Thus accumulations of water follow and evaporation from the surface of the soil concentrates the soil moisture near the immediate surface of the ground. The salt either crystallizes out onto the surface of the ground or diffuses down into the soil slowly. If this state of affairs is allowed to go on unchecked the surface of the soil becomes too saline for agricultural plants to grow and the soil is abandoned to salt grass and other alkali vegetation. The accompanying table shows the conditions found to exist in such soils after unchecked evaporation from a wet soil. Soluble salt content of Boswell sandy loam after irrigation. Depth. 135. 142. 27. 54. 144. 47. Feet. Per cent. Per cent . Per cent. Per cent. Per cent. Per cent. 1 0.09 0.06 0.45 0.66 0.16 0.41 2 .15 .06 .18 .44 .13 .52 3 .28 .10 a. 30 .52 .12 .48 4 .26 .14 .15 ».47 .10 .42 5 .32 .18 .17 .33 a. 10 .39 6 7 34 18 17 .33 a. 35 .36 a Water encountered. 48 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. Samples 135, 142, 144 illustrate soils in which water is not standing near the surface. Here the original salt has been washed down and away through the subsoil. In the other samples water has accumu- lated in the subsoil, and at the time examined stood at the depths noted. Through fluctuation in the level of this water table, the stand- ing water at other seasons approaches much nearer the surface. A general relation is to be seen between the depth to standing water and the amount of alkali within the surface foot of soil. The amount of alkali within the surface foot is the controlling factor in the cultiva- tion of these soils; and, since agricultural plants refuse to grow with ^ of 1 per cent alkali, the evaporation from the surface has already damaged some of the soils referred to in the table. R08WBLL I.OAM. This soil differs from the Roswell sandy loam in having the clay near the surface. A typical section shows about 4 feet of loam under- laid by heavy loam and clay. • The Eoswell loam is represented on the soil map by an area in the center of the Eoswell sandy loam. This area is underlaid by impervious clay throughout, and, wherever it has been watered without proper drainage, water can usually be found at a depth of from 3 to 5 feet. The conditions of this soil with reference to alkali are nearly the same as in the Eoswell sandy loam, but wherever evaporation has been allowed to progress unchecked the surface soil is at present charged with alkali. In some places this has gone on to such an extent that the soil should be classed as an alkali flat. Soluble salt content of Roswell loam. Depth. 33. 3J. 39.a 41. 48. 40. Feet. Per cent. Per cent. Per cent. Per cent. Per cent. Per cent. 1 0.19 61.34 1.69 2.76 0.42 0.72 2 .15 .45 .60 .93 .28 .36 3 C.22 .31 .43 1.14 c.22 C.31 4 .35 .31 .32 1.04 .20 .31 5 .36 .33 (c) 1.25 .16 .34 6 .36 .29 cl.04 .14 .32 8 10 12 .72 .44 .32 a Never i rrigated. i/Wat er at surfa ce. c Water er countered Sample No. 33 represents the soil in nearly its best condition as regards alkali, but with standing water too near the surface. The texture of the Eoswell loam in the field was classed as heavier than the Eoswell sandy loam, though mechanical analyses of a few samples show very little difference. This again illustrates the difficulty of inter- preting the texture of a soil itom a mechanical analysis alone. The change in arrangement of the soil particles by the action of irrigating waters and the presence of the clay nearer the surface give the impres- Report No. 64, U. S. Dept. Agr, Plate II 33 O V,i V ROSWELL DISTEICT HONDO MEADOWS. 49 sion in the Held that the Eoswell loam is a heavier soil than the Eos- well sandy loam or Pecos sandy loam, and it has been so classed. This shows that the holding power and its penetrability to water are often important factors iu determining the character of a soil as well as the size of the grains. THE HONDO MEADOWS. Under the title of Hondo Meadows are included the low-lying lands along all of the streams, but, with the exception of a narrow strip along part of the South Spring River and a small part immediately along the Pecos, these meadow lauds on the soil map are all along the Hondo River and around its junction with the Pecos. The term meadow is used iu the same sense as in humid regions, and means low land, naturally wet from the proximity of the stream, though in the arid climate of the Pecos Valley the low-lying lands need not necessarily be wet before irrigation. The soil of these meadows is formed from recent alluvium, and in its physical properties represents this mud, being heavy and silty. A typi- cal section of the Hondo meadow laud shows 3 feet of clay loam under- laid by clay. The water of the Hondo is very muddy, and when fields are flooded with this water a thin deposit is left. The fertilizing value of the silt in suspension is considerable; in fact an analysis shows it to contain more plant food than the mud of the ISTile River in Egypt. Chemical compositwn of Hondo and Nile sediment. Hondo mud (Skeats). ^Nile mud (Macken- zie). Insoluble matter and silica . Iron oxide and alumina Oxide of manganese Magnesia Lime Potash Soda Sulphuric acid Phosphoric acid Carbonic acid Organic matter Nitrogen in organic matter. 43.6 21.4 2.1 5.7 1.19 .32 58.17 24.75 .09 2.42 3.31 .68 .62 .20 .21 1.55 8.00 .12 The application of water containing this mud can not but be of bene- fit to most lands in the valley. Its wealth of plant food is above that of the ordinary soils, and the nitrogen it contains should be of special value. The sediment often amounts to as much as 10 per cent by volume of the water, it is stated. This analysis may be taken to represent the composition of the Hondo meadow soil. This soil is the richest in plant food of all the soils mapped. The sediment of the river represents the richest portion of the soils of the upper valley. H. Doc. 399 4 50 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. From the position of these lands, subject to weathering from below by seepage waters from the higher lands, the evaporation from the sur- face of the soil has accumulated quantities of salt within the soil and at present nearly the whole of this rich land is given over to salt grass. Cultivation has been attempted at several points and the efl'ect of underdrainage on the soils is readily noticeable. On section 33, one- half mile east of Eoswell, a truck patch has been laid out on the Hondo bottom. Near the bluffs sloping from the second terrace to the bottoms the land is wet and alkaline and the growth of tender truck crops uncertain or impossible. Toward the bank of the Hondo, which at this point is cut down 8 feet below the level of the meadows, the soil is better drained and truck crops do well. Should drains be cut from the river back to the bluffs at intervals, the water seeping out from the up- lands would be drained away and the salt already in the bottoms could be washed into the Hondo and so carried away. On section 34 an alfalfa field is growing on the Hondo flats. Along the foot of the bluffs, 150 yards from the river, the soil is wet and boggy, with salt grass the only vegetation. Close by the stream banks, however, the growth of alfalfa is heavy and luxuriant and grades down until, less than 100 yards from the escarpment of the terrace, it fails entirely to grow. Perhaps the best illustration of the effect of drainage is seen upon the farm of Mr. Charles Bremond, on section 31. Here the Hondo makes a bend, and close to the river all around the bend the soil is in far better condition than farther back. The mechanical analysis of the Hondo bottom land shows it to be a much heavier soil than any other exposed in the valley. The sediment as originally deposited contains very little more clay than the Eoswell loams, but from its position, being wet nearly all of the time, the soil particles break down and a much heavier soil is formed. The two analyses in the table represent the two conditions of the soil — the first column representing the sediment immediately after its deposition and the other soil formed from the weathering of this sediment. Mechanical analyses of soils of the Hondo meadows. Diameter. Conventional names. 4146. EosweU, 3 miles £. 4119. Eoswell, 1 mile E. MiUimeters. 2 to! 1 toO.5 0.5 to .25 .25 to .1 .1 to .05 .05 to .01 .01 to .005 . 005 to . f 001 G- ravel Per cent. Per cent. Trace. L46 5.58 21.04 35.69 7.13 14.52 Medium sand 0.76 4.85 10.65 12.35 17 34.65 Very fine sand. Silt Fine silt Clay vedin mechanical analysis. 0" C Salt dissol Loss at 11 .95 5.57 7.89 5.40 7.47 8.63 Loss on ig nition ROSWELL DISTRICT WATER SUPPLY. 51 Figure 1 gives sections through Roswell, Oarlsbad, Otis, and Florence and shows the relation of the soil at the different parts of the valley. SECTION THROUGH ROSWELL HB®^^-^^^^^^^^^^^— Ssr s ^_^^ i^^^^^^B ^B ^ ^ ^M kb3^^&^^ SECTION THROUGH CARLSBAD -CANAL 33<. _Conff! " _S PECOS R SECTION THROUGH OTIS SECTION THROUGH FLORENCE Fig. 1. — Sectione in Pecoa Valley, New Mexico. (S^^sand; .S'«c ^^ sandy loam ; *S'c=:loam; *%c = clay loam ; O^^clay; Oonj?^ ^^ conglomerate; Gyp^ gypsum.) Water Supply, The water supply from the North and South Spring rivers is good ; that is to say, it does not contain enough salt to be harmful to vege- tation in any way. The average water contains 75 parts of soluble matter to every 100,000 parts of water. About 49 jiarts consist of cal- cium carbonate (limestone) and calcium sulphate (gypsum), which are harmless to plants and crystallize out upon evaporation of the water. The remaining 26 parts consist of salts readily soluble in water and such as are likely to remain in solution and accumulate to such a degree as to be harmful. The following analysis by Prof. E. M. Skeats, of Oarlsbad, represents the average chemical composition of the salt dissolved in the water: Chemical composition of the North and South Spring rivers, in x)arts, per 100,000 parts of water. NaCl 8 CaCO:, - 18 CaSO, 20 MgS04 16 K2SO4 - 2 Silica, alumina, and iron 1 Water of crystallization, etc 10 Total solids 75 52 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. Since most of our agricultural plants are able to grow with their roots in solutions containing as much as 1 per cent of these readily soluble salts, it would require evaporation and concentration of the Spring Eiver water to one-fifteenth of its bulk in order to kill vege- tation. Such a concentration can easily be avoided within the soil. By the application of 1 acre-foot of this water about 700 pounds of harmful salts are added, and, since 15,000 pounds per acre-foot can be taken as the maximum amount of salt allowable for plant growth, it would require nine years' accumulation of irrigation waters, allowing 2^ acre-feet of water for each year, and supposing none of the salt washed out of the land. Such a condition could hardly exist, for a certain percentage of the salts is washed down below the first foot. Moreover, part of the salt is annually removed in the crop. Four- tenths of 1 per cent of soluble salt in a soil capable of holding, when saturated, 40 per cent by weight of water gives a concentration of 1 per cent in the soil moisture when the soil is saturated. Underground Water. One of the maps accompanying this report shows the depth to standing water during June, 1899. The depth is shown by three shades of green, the lightest shade showing land in which water can not be found by boring 10 feet. The intermediate shade shows land in which water is to be found between 3 and 10 feet in depth. Such land as this, if the water approaches the higher limit, is in danger of becoming too wet and the level of the water table should be carefully watched ; and, if it rapidly approaches 3 feet, drainage must be fur- nished to prevent the water from rising above 3 feet. The reason for giving this laud a special tint is that it is approaching the limit, and may need drainage. The darkest tint shows land which is at present in need of drainage, for at the time it was examined water was found within 3 feet of the surface. For such land there can be no question but that its great need is underdrainage. Few plants, none of which are used as agricultural plants in the Pecos Valley, are able to grow to advantage with their roots immersed in water for more than a day or two at a time, and, as most of our common plants send roots as deep as 3 feet, the level of standing water should be kept below this level. Particularly is this true of arid regions, where there is not so much difference between soil and subsoil as in the East. Alfalfa, which is the most important crop of the Roswell area, sends its roots to great depths. Cases have been noted where the alfalfa grew luxuriantly for two or three years, but suddenly began to sicken and die. Investi- gation proved that the plants grew all right until the roots reached the water table. Here, in their effort to reach farther into the subsoil, the roots were partly immersed in standing water and the crop suffered. The scale upon which this map is printed does not permit detailed representations of each section of land. There are no doubt small spots KOSWELL DISTRICT ALKALI OP THE SOILS. 63 within the area shown as in need of immediate drainage which are at present in fair condition for crop production. The map is intended to show in a general way the conditions found existing duriug June, 1899, and must not be interpreted as representing the conditions at any other time. The entire condition of the land may change with the seasons, and during some other season of the year the water level may be very diftierent. On the map all of the lands below the irrigating canals have standing water at less than ten feet. Such is generally true, though there are a few isolated points immediately along the bank of the Hondo, or under some land never irrigated where water stands at a greater depth than ten feet. Throughout the greater part of the Eos- well loam, where clay is found near the surface of the soil and in the draws and close to the river channels, the land is wet and water stands at less than three feet. The Alkali of the Soils. One of the maps accompanying this report shows the conditions of the ground with reference to alkali. The three colors indicate the amount of alkali in the soil; soil containing from to 0.25 per cent of alkali; soil containing 0.25 to 0.50 per cent, and soil containing over 0.50 per cent. Or the map shows (1) the areas of land which are free from harmful quantities of alkali; (2) the lands which contain harmful quantities of alkali, but not enough alkali to prevent the growing of crops; and (3) the lands which contain too much alkali for cropping. The term alkali, as used throughout the West, designates any solu- ble matter in a soil or water. The term does not mean that the salts are alkaline or basic in a chemical sense, for the greater part of these salts are neutral, and in some cases even acid salts have received this name. There are two kinds of " alkali " commonly recognized, [a) white alkali and (&) black alkali. The white alkali is composed of one or more of the following salts, named in the order of their importance : Sodium chlorid, sodium sul- phate, magnesium sulphate, calcium chlorid, magnesium chlorid with, in some districts, borates and nitrates. The essential constituent of black alkali is sodium carbonate, though this salt is never pure, but is mixed with sodium bicarbonate and the salts which form the white alkali. These salts are formed through the weathering of the rocks, and in dis- tricts of light rainfall the amount of seepage through the soil is insuffi- cient to wash them away. There is therefore an accumulation in most arid soils — the amount, other things being equal, varying inversely as the rainfall of the district. In some places the alkali is the accumulated salt from the evapora- tion of inland lakes. All river water contains soluble matter, and when this water enters an inclosed basin where the waters evaporate an accumulation of salts is the result. The muds, silts, and sand left by this lake all contain soluble matter, and upon applying water by irriga- tion to such soils the alkali, which may not be apparent on the surface 54 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. of the soil, is brought to the surface by the water, as it rises through capillary attraction. These inclosed basins sometimes represent the most difficult conditions with which the farmer has to deal, for the salt commences to accumulate in the lowest portions of the basin and extends up the sides of the basin as the amount of salt brought in increases. The great expense or even impossibility of draining such basins is the great barrier in the way of their reclamation. The Great Salt Lake of Utah is the remnant of a great lake, and its waters to-day represent the mother liquor from the concentration of great quantities of water. In past geological ages arms of the ocean were sometimes cut off from the main ocean and the waters concentrated by evaporation until part or all of the burden of soluble matter was deposited. Beds formed in this way to-day, when elevated above the level of the sea and exposed to erosion, give rise to quantities of alkali. The Ked Beds, forming the great plains north of Eoswell in the Pecos Valley, and underlying the Staked Plains to the east, were deposited in this way, and though no beds of salt are found, yet all of the soils from this for- mation contain soluble matter. Beds of gypsum were deposited, thus indicating that the ocean water was concentrated by evaporation. Included within this gypsum are small quantities of soluble salt. If beds of these salts were ever deposited, the greater part has been removed by solution, for to-day no deposits of salt within this forma- tion in New Mexico have been found, so far as known.' Another source of alkali in the Pecos Valley is from the decomposi- tion of volcanic rocks. To the west of the Pecos great areas of appar- ently recent lava are found, and associated with the lava are alkali springs of varying degrees of concentration. Some of the deposits of gypsum in the Pecos Valley can be accounted for only on the assump- tion that the gypsum is the result of the action of acid waters upon the limestone. Even though the lava flows were not associated with alkali springs, the decomposition of the igneous rocks would give rise to the formation of alkali salts, and without doubt the igneous rocks west of Eoswell are the source of parts of the alkali salts. The alkali of the Roswell area is entirely the white alkali, as far as observed. Many spots were pointed out as being due to the black alkali, but they were undoubtedly due to calcium chlorid. Sodium carbonate can hardly exist under the prevailing conditions. It has been pointed out by Hilgard that gypsum is an antidote for black alkali under certain favorable conditions. If the soil be well drained and aerated, the black alkali, sodium carbonate, in contact with the gypsum is more or less completely converted into the less noxious white alkali, sodium sulphate. The reaction may be expressed thus : Na2C03-|-CaSO4=Na3SO4-fCa0O3 ' Since this was written it is reported tliat a bed of rock salt over 400 feet deep has been encountered in boring for an artesian well at Carlsbad. ROSWELL DISTRICT ALKALI OF THE SOILS. 55 On the other band, if the soil be not well drained, that is to say, condi- tions exist favorable to the retention of carbon dioxid and moisture, a reverse reaction will predominate, which may be indicated thus : Na«SO4+0a0O3 + 0O2+H2O=Na.iSO4+H20a (003)2=2 H NaC03+ CaS04 and the sodium bicarbonate (H NaCOs) being a very instable com- pound, always inverts to some extent with the formation of sodium carbonate (Naj CO3), the undesirable constituent of black alkali. Somewhat similar reactions take place when gypsum comes in con- tact with chlorids in solution or in a wet soil. Calcium chlorid and sodium sulphate will be formed, which is indicated thus : 2 ]sra01+CaSO4=]Sra2SO4+GaCl2 There will now be present in the solution, sodium chlorid, calcium sul- phate, sodium sulphate, calcium chlorid, as well as the ions formed by the dissociation of these salts. If the soil be well drained and suffi- cient water drains through it, the soluble salts thus formed are rapidly leached away ; but if the soil be not well drained, that is to say, if it contains standing water near the surface, there will be a gradual evaporation of the water with a concentration of its salt contents. At first sight it would appear that the calcium sulphate, being so much less soluble than the other salts, would be precipitated as a solid before any of the other salts ; more and more being formed from the sodium sul- phate and calcium chlorid and precipitating in turn. But this process will take place, if at all, to a much less extent than is generally sup- posed. It is to be remembered that the solution is becoming more and more saturated with respect to the sodium chlorid. It follows, from the investigations of Treadwell and Eeuter, and from experiments in this laboratory, which will be described later by Dr. Cameron, that the amount of gypsum dissolved — its constituents being held in solution as calcium chlorid and sodium sulphate — is very greatly increased with the concentration of the sodium chlorid. It thus comes about that as evaporation proceeds the solution is actually becoming richer and richer in the very soluble calcium chlorid, and finally it separates as such along with or after the other readily soluble salts. On resolution, being much the most soluble of the salts in the mixture, it would be the first to dissolve, and from the fact that it has a common ion with sodium chlorid, would retard the solution of that otherwise very readily soluble salt. It would increase the solu- bility of the less readily soluble sodium sulphate if it remained in con- tact with it sufficiently long for equilibrium to take place. But if it is quickly brought to the surface by capillary forces, or by diffusion, and then quickly concentrated by rapid evaporation, there would be a con- centration of the calcium chlorid. in the upper portions of the crust. 56 A SOIL SURA^EY IN THE PECOS VA.LLEY, NEW MEXICO. In several spots in the Pecos Valley considerable quantities of cal- cium and magnesium chlorid were found in the crust on the soil. It is probable, as will be seen from the above reasoning, that but small quantities would be found at any dejith in the soil itself. These crusts containing calcium chlorid have a dark appearance quite similar to the true black alkali. Indeed, they are locally known as black alkali, though their characteristic feature is the presence of much calcium or magnesium chlorid, not that of sodium carbonate. The reasons for this blackened appearance are not obvious, and the evidence at hand is too meager to warrant any suggestion at present. It is worthy of note that these spots are of necessity damp or wet — a condition favor- able to the formation and retention of carbonic acid. But it will require further investigation to say definitely whether this be a factor in the production of the observed results. Analyses of alkali crusts from the Boswell area. No. Locality. Calcium sulphate. MagueBiuiu sulpbate. Sodium sulpbate. Sodium chlorid. 4092 4093 4097 4107 4139 4149 Sec. 35, T. lOS., E. 28E Sec. 35, foot of gravel hill Bremoud crust Black crust, Bremond High spot, Michelet ." Tule crust, E. of Northern canal Per cent. 14.17 9.09 13.58 6.85 45.57 4.03 Per cent. 28.05 38.42 16.73 lZ.86 17.53 21.64 Per cent. 47.99 41.15 35.77 50.90 1.44 64.77 Per cent. 9.82 11.33 34.90 29.41 35.46 9.57 Of the salts shown in these analyses sodium chlorid is the most harmful to plants. It is generally stated that most agricultural plants can grow with 0.25 per cent within the surface foot of the soil, while of the sodium and magnesium sulphate plants can withstand as much as 0.50 per cent. Observations upon the growth of alfalfa in the Eoswell soils showed that it was just able to grow when from 0.40 to 0.50 per cent of alkali was present. The figure 0.60 has been taken as the maximum limit, and though plants are able to grow with this amount of salt in a soil the growth is uncertain and light, in fact 0.25 per cent damages crop growth to a great extent. This range is, there- fore, taken as the danger limit. The action of soluble matter within a soil upon the growth of plants is very complex and to a great extent not well understood. It is a fact of common note that a little alkali makes crops grow better. There is danger in this little alkali, however, for if the amount increases ever so little beyond the point where plants grow better, there is noted a decrease in production and the productivity of these soils may be considered dependent upon the per cent of alkali present. The fact has been noted that with 0.25 per cent of alkali in the top foot of soil the crop pro- duction is decreased. This decrease becomes more and more marked until the plants refuse to grow. The limit of growth may be considered 0.50 per cent. Occasional cases have been noted elsewhere where crops would live with more alkali present, but in the Pecos Valley, soils, with one-half of 1 per cent alkali, are entirely unproductive. ROSWELL DISTRICT ALKALI OF THE SOILS. 57 Plants of different species are able to withstand different amounts of alkali, and the resistance of the same plant to alkali varies in different stages of its growth. In the Roswell area afalfa was found sickly and dying with a little over 0.40 per cent alkali in the soil, while in other stages of its growth isolated plants have been found growing in a soil containing more than 0.75 per cent alkali. Young apple trees showed signs of distress with a little over 0.30 per cent, while sugar beets grew with something over 0.50 per cent. The action of soluble matter upon plant growth varies with the osmotic pressure of the solution in the plant tissues. This fact has been experimentally demonstrated by Slosson and Buffum of the Wyoming Experiment Station. Since the osmotic pressure varies with the concentration of the solution, the mere statement of the total percentage of salt within a soil determined chemically does not give a true idea of the osmotic pressure of a solution. The percentages as given in this report and in all reports of the Division of Soils are based upon the dry weight of the soil and are the amount of alkali dissolved by the water when the soil is saturated. The method for salt determinations as used in the fiield work of the Division of Soils has been described in previous publications, but a short sketch of the method will be given here for the intelligent con- sideration of the results. Samples of soil ai-e collected with a IJ-inch soil auger with an exten- sion handle long enough to bore 20 feet. The samples are usually taken in 1-foot sections and the alkali determined in each foot of depth. The soil as collected is thoroughly mixed upon a clean piece of oil- cloth. A portion of the mixed sample is saturated with distilled water and packed in a hard-rubber cell of a capacity of about 50 c. c. with two metal sides. The electrical resistance of this saturated soil is taken by a Wheatstoue bridge. The temperature is then taken by inserting a thermometer in the soil in the cell. This constitutes the field work. In the laboratory a number of typical samples of the alkali in the soil are analyzed and a solution of salt corresponding in relative composition to the average of these analyses is made. By varying dilutions of this solution the per cent of salt is ascertained for each 5 or 10 ohms difference in resistance. The pei^ cent of water of saturation for each type of soil is then determined by saturating the soils in the same manner as done in the field and making gravimetric moisture determinations. Then by determining the weight of the wet soil the cell will hold, the number of cubic centimeters of the soil moisture in the cell filled with saturated soil may be calculated. Knowing the number of cubic centimeters of soil moisture contained in the cell full of saturated soil, the cell factor, or the figure by which the resistance of that amount of solution must be multiplied in order to give the specific resistance, can be found from the known constants of the cell. The resistance obtained in the field is multiplied by the figure 0.55 to correct for the resistance offered by the insoluble, nonconducting parti- cles of soil and the result multiplied by the cell factor, giving the 58 A. SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. specific electrical resistance of the solution around the soil grains. The per cent of salt corresponding to the specific resistance is deter- mined from the solution made up from the chemical analyses of the field samples. This percentage, multiplied by the per cent of water in the saturated soil, gives the per cent of alkali in the dry soil. In the field work borings are taken at frequent intervals and the salt con tent determined for each foot of depth. These percentages are plotted on base maps and lines drawn separating the three conditions of the soil respecting alkali. These three conditions are shown in the alkali map by different colors. The maps as presented may be taken as an indication of the condition of the soil during June, 1899. For the construction of the maps the salt content of the soil to a depth of 5 feet was considered. The area of Pecos loam is seen to be colored to indicate between 0.25 and 0.50 per cent alkali. This soil at every point examined within the area of the map was found to contain more than 0.25 per cent of alkali within the upper 5 feet. The porosity of this soil, however, would enable most of this salt to be immediately washed out upon the appli- cation of irrigation water, as has been seen in a number of places where small tracts of this upland soil were cultivated around Eoswell. The areas on the salt map showing from 0.25 to 0.5 per cent of salt are susceptible of a much greater range of conditions than are indi- cated by the range in salt content to the depth of 5 feet. Crops may make a moderate growth on any or all parts of this class of land, and there are places near the lower limit, undoubtedly, where the con- ditions are excellent at the present time, because the salts are so dis- tributed that only small amounts are contained in the first and second foot, while a sufficient amount occurs below this to bring the average up to 0.25 per cent or more. This condition, while giving excellent results at present, is dangerous, because the salts are liable to move toward the surface unless good underdrainage is provided. As we approach the upper limit on this class of land the other extreme may be found to exist, i. e., an accumulation of salts at the surface in sufiflcient amount to kill all crops, and yet the quantity of salts be suflicieutly small in the lower depths to bring the average for 5 feet below 0.5 per cent. The Problems ov the Eoswkli. Area. It has been shown under the discussion of the soils that the accumu- lation of the alkali salts in the surface of the soil is in all cases due to lack of drainage. The salts which are present now near the surface were once buried at such a depth as to be of little or no damage to plants, and this translocation has been brought about by irrigation. The water supply is good; therefore the salt must originally have come from the soil. Much of the land has already suffered from accumula- HAQERMAN AREA — GEOLOGY AND SOILS. 59 tious of salt, so there are two important problems before the farmer: First, the prevention of further damage from seepage water and alkali salts; second, the reclamation of the lands already damaged or abandoned. Eoswell is situated in a large grazing country, and the principal industry is cattle and sheep raising. The agriculture is largely sup- plementary to the range. Alfalfa is the principal crop. There is, how- ever, a growing tendency toward fruit raising on the part of some of the farmers. The local demand for fruit is great enough at present to use all of the fruit produced, but there is danger of overstocking the local market, and thus necessitating the shipment to distant markets. Situated as it is, there is no question but that cattle and sheep will prove the most profitable industry toward which the farmers may turn. The success of the Eoswell community is to be attributed to their realization of this fact. Since alfalfa offers the most ready forage crop, the soil should be in the best condition for its growth. One of the most essential conditions of a soil in an alfalfa field is good drainage. Alfalfa roots penetrate deep into the subsoil, and as soon as standing water is encountered their roots cease to grow. If drainage keeps the level of standing water down, the success of the crop is insured. Drain- age also removes and insures }io further damage from alkali. THE HAGEEMAN AREA. GEOLOGY AND TOPOGRAPHY. The Hagermau area, or Northern Canal system, lies about the center of the basin formed by the ancient obstructions in the Pecos River about Seven Eivers. The Eoswell area lies near the northern end of the same basin. This basin, cut out of the Eed Beds or Permian strata, is more or less filled with sediment derived from these rocks. The Eed Beds contain crystals of gypsum scattered throughout their sands and gravels, and, interstratified with the beds of sand and clay, heavy, massive saccharoidal gypsum. It is to the solution and redeposition of this gypsum that the gypsum of the soils of the Eoswell area is mainly due, though this gypsum in the soils may in part be from mechanical sedimentation. The Hagerman area lies along the terraces cut by the Pecos Eiver in the sediments of the old lake. The action of the Pecos has been aided by the streams entering from the west, notably by the Felix and Penasco. The terraces are very imperfect, the land sloping from the edge of the river channel gradually upward to the west. On the east side the Pecos is bounded by high bluffs. SOILS. The soils of the Hagerman district correspond very nearly to the soils of the Eoswell area. On the lowest lands lying close to the Pecos Eiver dune sands and sandy loams form the predominating soil types. These are usually well drained from their nearness to the river chan- nel, though occasionally, even when apparently in the most favorable 60 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. coiiditious, these lands are found wet and in need of drainage. On the level lands immediately around Hagerman, the soil is heavy and corre- sponds very nearly to the Koswell loam. Around this are areas of gypsum soil, which is exposed in drains to the south of Hagerman. The discussion of the cultivation of this gypsum land will be deferred until it is considered under the Carlsbad area, where it is more typi- cally developed. Upon the uplands of the Hagerman area and above the canal the Pecos loam, similar to the upland soil at Roswell, is typically devel- oped. This area of soil is a continuation of the Roswell area. A soil map has not been drawn of the Hagerman area, as a short time only was spent in reconnoissance of the soils. The great similarity of this district to the Roswell and Carlsbad soils, however, will enable the conclusions drawn from them to be readily applied to the Hager- man conditions. THE WATER SUPPLY. The Northern Canal heads in a dam across the Hondo River directly east of Roswell. This dam collects and directs into the canal the unused and waste water from the Berendo and North Spring rivers, also from the Hondo in times of flowing in that stream above Roswell. A great deal of seepage and drainage water is collected, which, together with the salt from the Berendo water, renders the Northern Canal water salty. At the point where the Northern Canal crosses the South Spring River a second diverting dam collects the water of that stream. This also contains quantities of seepage and drainage water. The average condition of the Northern Canal water is shown by the following anal- ysis by Prof. E. M. Skeats, of Carlsbad, N. Mex. Parte per 100,000. NaCl 65 CaCOs 17 CaS04 51 MgS04 25 K2SO4 ;.l 23 Na,SO ' ""■ Silica, alumina, and iron 2 Water of crystallization 19 Total solids 202 Of these salts calcium carbonate, calcium sulphate, the silicon, iron, and alumina compounds, with much of the water of crystallization, amounting in all to 89 parts, may be disregarded, since upon the evaporation of the water these compounds crystallize out and do not collect in the soil in sufficient quantity to be of harm to vegetation. Thus the harmful salts of the water are present in the i^roportion of 113 parts of salt to 100,000 of water. UNDERGROUKD WATER. The wells above the canal are from UO to 100 feet deep, with water at about 60 feet. The water in all of these wells is good, containing about ROSWELL DISTRICT PROBLEMS. 61 35 parts of soluble matter per 100,000. Of the land below the canal, part of it has already become so wet that underdrainage is necessary. Around the town of Hagerman water is found in places as near the surface as 3 feet. Daring the four years during which irrigation has been practiced the entire character of the vegetation around Hager- man has been changed from grama and other prairie grasses to salt grass. The change has been brought about almost entirely by under- ground seepage waters and alkali. ALKALI OF THE SOILS. The condition of the soils of the Hagerman area correspond exactly to the conditions existing at Eoswell. The cause of the present condi- tions, however, includes one factor which was noted as unimportant in the Roswell area; that is, the amount of alkali in the irrigation water. The application of each acre-foot of Northern Canal water adds 3,000 pounds of salts to the soil. Such salts, if allowed to accumulate by evaporation of the water, would soon so impregnate the soil with alkali that crops would not grow. Allowing 2^ acre-feet per year, in two years 15,000 pounds of soluble matter would have accumulated, or enough to prevent profitable cultivation, provided all of the salt was retained by the surface foot of soil. This rapid accumulation may easily be pre- vented by washing the accumulated salts from the previous irrigation down into underground drainage. Another successful method is to grow soil-shading crops which prevent the rapid evaporation from the surface of the soil, so that less water is needed in irrigation. Cultiva- tion offers the most effective method of preventing evaporation from the soil's surface, and this should be carefully followed out in the growth of all crops which allow cultivation. PROBLEMS OF THE DISTRICT. The problems of the Hagerman district are very similar in character to the problems of the Eoswell district. The accumulation of alkali salts must be prevented, and the excess already present must be removed. The leakage from canals and laterals, particularly in the case of water containing alkali salts, is one of the most potent sources of trouble. The loss from the canal in running through the sandy and gravelly soils is great. In the case of the main canal 1 cubic foot per second is lost for every mile. This water, seeping under the ground, gradually fills up the subsoil until capillary force lifts the water to the surface, where it evaporates leaving the alkali. CAKLSBAD AREA. Geology. The Carlsbad area lies entirely in Eddy County, New Mexico, along both sides of the Pecos River. The irrigation district lies in a basin cut off on the north by a canyon, through which the Pecos runs, and on 62 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. the south by canyons and rough country from below the Black Elver to the Texas line. This basin is undoubtedly the site of an ancient lake, bounded on the west by the foothills of the Guadalupe Mountains, on the east by the scarp of the Staked Plains. The entrance to and exit from this basin were through narrow gorges cut in the hard carbonifer- ous rock of the foothills. The sediments form the present soils of the district, modified to a great degree by the Pecos and the drifting action of the wind. The elevation of the country, or more likely the cutting away of the dam below Black Eiver, placed the Pecos in position to commence the removal of the lake sediments. The river has begun the cutting and has already worn down to its hard rock over part of the district, and has commenced meandering in its bed, forming broad and treacherous sand bars. At other points in its course the Pecos has worn down to the conglomerate which lies at the base of the lake sediments. The water supply of the Carlsbad area is derived entirely from the Pecos Eiver. The original plans for irrigation contemplated taking the water direct from the river with a storage reservoir 6 miles above Carls- bad, with a capacity of 7,000 acre-feet. The situation of the dam is in a canyon with limestone walls, offering a natural abutment for the ends of the dam. This storage reservoir proving inadequate, a second reser- voir was constructed 17 miles above Carlsbad, near Seven Elvers. This dam forms a lake 8 miles long and of an average width of If miles, with a storage capacity of 140,000 acre-feet. Soils. Map 4 accompanying this report shows the soils of the Carlsbad area. Four distinct types of agricultural soil have been recognized and named, each type being different in its relation to irrigation waters and alkali and to vegetation. These types are as follows: 1. Pecos sands. 2. Pecos sandy loams. 3. Pecos conglomerate soil. 4. Gypsum loam or " yeso." THE PECOS SANDS. Along the Pecos Eiver, filling up the tortuous bends of its course, are areas of sands blown about by the winds. The characteristic vegeta- tion of mesquite, yucca, and canaigre in a great measure prevent the drifting of these sand dunes. The mesquite forms small dunes around its roots, and as the dunes build up the mesquite extends its roots, keeping on top of the dunes. In this way dunes 8 to 10 feet high are formed with mesquite growing on the top. This dune sand is composed of rounded grains of quartz with smaller quantities of gypsum and lime- stone. The texture of the dune sand is shown in the following table: Report No. 64, U. S Oept, Agr. Plate III. 05 > = o 1 o CARLSBAD AREA PECOS SANDY LOAM. Meclianical analysis •of dune sand. Diameter. ConTentional names. 4064. Carlsbad. 2 miles NW., Otoe inches. Millimeter. 2 tol 1 toO.5 0.5 to .25 .25 to .1 .1 to .05 .05 to .01 .01 to .005 .005 to .0001 Per cent. 0.00 .00 .22 70.25 21.50 2.70 .43 4.22 Fin© sand Very fine sand Silt Fine silt Clay w n .43 .35 This type of sandy soil is from its texture well adapted to truck farming and root crops. Melons, potatoes, and small fruit grow well on this soil. It forms the soil of La Huerta, that part of Carlsbad north of the Pecos River and immediately adjoining the town. There are narrow strips of this soil found along the river as far south as the area was examined. Such a soil is not apt to be rich in plant food, but in the forcing of fruit and early vegetables this is not so necessary as ease of cultivation. The virgin Pecos sand contains very little alkali. Examinations were made at points all over the area of the portion mapped out, and in no case was natural accumulation of alkali found. The alkali profile, in Figure 2a, shows the per cent of alkali in two typical soils, one never irrigated and the other irrigated. It will be seen that the irri- gated soil on the average carries a little more alkali than the virgin soil. THE PECOS SANDY LOAM. This is the most important type of soil in the Carlsbad area and covers the greater part of the irrigable land under the Southern Canal. This loam is very similar in character and texture to the Pecos sandy loam at Roswell. The accompanying table shows the average texture of the soil. Mechanical analyses of Pecos sandy loam soils. Diameter. MUlimeterB. 2 1 0.5 25 .1 .05 .01 tol too. 5 .25 .1 .05 .01 .050 .005 to .0001 ConTentional names. Grravel Coarse sand Medium sand... Fine sand... Very fine sand. Silt Fine silt Clay Salts dissolved in 1 J liters of water, nsed in mechanical analyses Loss at 110° C Loss on ignition 4067. i mile S. Otis, N. Mex., 0tol2 inches . Per cent. 0.00 Trace. 1.32 10.20 36.30 19.10 3 18.65 4 2.60 4.70 4068. IJ miles S. Otis, N. Mex., 0tol2 inches. Per cent. 0.00 Trace. 1.65 12 41.45 10.27 10.45 18.72 2 2.67 2.45 64 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. The mechanical analyses of 4067 and 4068 represent typical soil from the Pecos sandy loam. These soils in the field appear much lighter in texture than the mechanical analysis would indicate. A microscopical examination of the soils was made to see the cause of this. Under the microscope a great many particles were found to consist of little con- glomerates of clay cemented by lime. These particles upon shaking in the mechanical analysis gradually disintegrated, forming a soil which contained 32 per cent of clay. Since a mechanical analysis made in ALKALI PROFILE 0-/ .^r^j/. *•/. FT 1 ALKALI PROFILE o'A i% .z%2% A°A .&% .e% y% st/' .h'/- loX fc3 ^ \ 'i 1 \ \ \ .^•'' 1 \ K ^ ( / \ \ i / a. In Pecos sand. Before irrigation After irrigation. Tig. 2.- b. In Pecos sandy loam. Original soil. Irrigated soil. subirrigated soil Diagram showing salt content to depth of 6 feet. this way did not represent true field conditions analyses were made with only enough shaking with water to wash off the loose clayey par- ticles. Also in soils containing gypsum and lime, the water used for the mechanical analysis dissolved a quantity of the gypsum and lime. This amount is indicated in the table as soluble in water, but must not be taken to represent the per cent of alkali salts. In the field the breaking down of the particles is observed in spots continuously wet, and even with a normal amount of moisture the soils have become a little heavier by this disintegration. In some parts of the area this disintegration of the soil has gone on since irrigation com- menced, and now we have irrigated soils which are heavier than non- irrigated soils situated under the same conditions. The alkali profile given in figure 2b shows the per cent of alkali salt in the soil under three conditions: (1) The virgin soil; (2) the soil after irrigation, where drainage is good; and (3) the soil where irri- gated when the drainage is poor or when subirrigated. The texture profile on the soil map shows the soil to be composed of nearly all (93 per cent) sand to a depth of 6 feet. Situated as these sands are, along the banks of the river, the drainage is normally good and the salt, which originally is found in the subsoil, is readily leached out and. CARLSBAD AREA — PECOS CONGLOMERATE SOIL. 65 carried away. Where this draiuage ia defective the evaporation from the surface is sufficient to rapidly coucentrate the soil solutions and the alkaili salts accumulate at the surface. Such a state of affairs is found along the foot of the gravel bluffs, in sees. 8, 9, 10, 14, and 15, of T. 22 S., R. 27 E. (See map 5.) Here the water in seeping out from the conglomerate bluffs has swamped an area of land and the rapid evaporation of the water is swiftly producing an alkali flat. The Pecos sandy loam contains about 20 per cent of carbonate of lime. So much lime is objectionable to a great many plants, but fortu- nately the class of plants grown in the, Pecos Valley are tolerant of lime. The humus content of the soil is low, as is true of many of the arid climate spoils. Since the humus is the principal element of nitro- gen, this element is perhaps the element most needed in the Pecos Valley soils. The plowing under of alfalfa is found to be of great benefit to sugar beets, due to the addition of the nitrogenous organic matter of the alfalfa. The practice of plowing under alfalfa has not been general, since the cost of seeding and starting alfalfa is heavy. The iirst year's crop is not sufficient to pay for the cost of planting. There are other leguminous crops which, without doubt, could be grown in the valley, and could be turned under with profit. The question of fertilizing, either with green manures or chemical fer- tilizers, is one which farmers of the West are very loath to consider. Western soils are usually rich soils, but in some districts even the virgin soils are poor. Particularly is this true of soils in the true arid regions, where the native vegetation is scanty and the humus content of the soils very low. The yield of sugar beets per acre upon the Pecos loam averages less than 5 tons per acre. This low yield is due partly to lack of ijroper plant food within the soil. No experiments, so far as known, have been conducted on the effect of fertilizers other than alfalfa upon the growth of beets or other crops. This defect in the soil is realized by the far:ners at present, and improvement in their methods of handling the soils is looked for in the near future. PECOS CONGLOMERATE SOIL. Underlying the Pecos sandy loam throughout the area examined is found a bed of gravel or conglomei'ate. This conglomerate is exposed over many points in the area, and whenever so exposed its disintegra- tion gives rise to a gravelly soil. This loose material is usually 2 feet deep, underlaid by gravel or conglomerate. The table following illus- trates the texture of the soil. No. 4069 contained 19.3 and 4070 con- tained 26.1 per cent of coarse gravel. H. Doc. 399 5 66 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. Mechanical analysis of Pecos conglomerate soil. "Diameter. Millimeters. 2 to 1 1 to 0. 05 0.05 to .25 .25 to .1 .1 to .05 to .01 to .005 to 05 01 005 0001 Conventional name. Fine gravel.... Coarse Band Medium sand.. Fine fland Very fine sand. Silt Fine silt Clay Loss at 110' C... Loss on ignition. Trace. Trace. 1.20 1.11 1.80 1.97 * 12.51 12.88 45.77 39.81 12.91 17.66 3.43 3.46 17.22 18.22 2.21 2.37 2.99 4.08 Presenting as it does such a large percentage of gravel, this soil is difi&cult to cultivate. Water readily leaches through it, and its irriga- tion would require large heads of water to cover short distances. Very little of this gravelly land has been cultivated. The area is seen to be large, and as land becomes scarcer and more valuable there will no doubt be a demand for this gravelly soil. The ready penetration of the water in this soil makes low-lying areas of it uncer- tain, for water seeping up from below through the con- glomerate soon swamps the overlying soil. The canal over part of its course runs through the con- glomerate. At these points quantities of water seep out of the bed and through the conglomerate to appear low- er down. The alkali flats iu several sections have been formed in this way from the seepage water out of the conglomerate bluff into the sandy soils along the river bottoms. This permeability of the conglomerate offers one of the great sources of dan- ger from seepage water. Fig. 3.— Diagram of orchard showing depth tostanding water. ThOUffh it offcrS a read V nat- ural underdrainage for part of the upland, it is in the same way a great source of damage to the low-lying lands. Draws and flats adjacent to V^ ^\ N? ^ \^ >^' \\\ Report No. 64, U. S. Dept. Agf. Plate IV CARLSBAD AREA GYPSUM LOAM. 67 the hills of conglomerate receive as seepage water the drainage from the upper lands. Figure 3 shows the depth to standing water in an orchard near Carls- bad being injured by seepage water and alkali coming from under the conglomerate. The arrow shows the direction of the underground drainage toward the Pecos Eiver. The surface of the orchard is nearly level and about 20 feet above the river. This shows the influence of the good drainage along the river bank, in lowering the level of the standing water. Figure 4 shows the soluble salt content of the surface foot in the orchard and the influ- ence of the good drainage along the Pecos Eiver. The relation of these lines to the drainage lines in figure 3 is very apparent. GYPSUM LOAM OR "YBSO.'' The soil map shows a large irregular-shaped area of gyp- sum soil north of Black Eiver and another area south of Black Eiver, around Malaga. This gypsum or "yeso," as the native Mexicans term it, possesses marked peculiari- ties which merit a careful study. The material is composed of nearly pure gypsum in a granular form. When dry it is Ql hard and compact, but | I 7 / K I ^ \^5 ^ ^\ upon wetting it absorbs water about as readily as a lump of sugar and breaks down into a soft mass which is very per- vious to water. Mechanical analyses of this material are very indefinite because the particles are so soft that they break up by agitation with water. The area mapped as gypsum soil does not everywhere have the gypsum on the immediate surface, but in nearly all places is covered with a thin layer of loam, derived from the decomposition of -the gypsum or deposition by wind and water of other material. This loam varies in thickness from a fraction of an inch to 3 feet in depth. Fia. i. — Diagram of orchard shotting soluble salt content of surfaoe foot. 68 A SOIL SURVEY IX THE PECOS VALLEY. NEW MEXICO. The cdtapositioii of the gypsum loam is shown in the accompanying table : Meclianical analyses of gijpsnm loam noils. Millimeters. 2 to! 1 too. 5 0.6 to .25 .25 to .1 to .05 to .01 to 1 05 01 005 .005 to .0001 Conventional names. Gravel Coarse sand . . . Medium sand.. Fine sand Very fine sand . Silt Fine silt Clay Salts dissolved in li liters of water used in mechanical analyses Loss at 110° C Loss on ignition Per cent Trace. 2. 2C 0.42 38.47 12.67 10.87 19.70 2.58 14.25 37.13 13.95 8.88 15.46 1.90 3.03 3.41 2.34 2.80 3.27 The most important physical property of the gypsum is its capillary power. When compared with another soil this property of the gypsum is remarkable. Wet spots have been observed in a field where on boring water could not be found within 6 ieet of the surface. A tube filled with the gypsum soil was placed with its lower end in water and the height to which the water rose was determined every day. The following table shows the results of the experiment: Time. 15 minutes 45 minutes 1 hour 2 hours . . - 6 hours . - - 24 hours . . 2 days 3 days 4 days Pecos Pecos gypsum, neight in dune sand. height in inches. inches. li 6i 2i 114 3 13 6 14 9i 17 17J 18i 24 19J 27i 20i 32 20J 6 days 7 days 8 days 9 days 10 days 13 days 14 days 15 days 25 days Pecos gypsum, height in inches. Pecos dune sand, height in inches. At the time the top of the tube was wet the water was slowly rising, and no doubt would have risen several inches higher. This great capillarity is of special importance and must be fully considered before reclamation by irrigation. The surface of the soil is kept moist all the time and the rapid evaporation from the surface causes the dei)osition of the alkali salts. It is undoubtedly this physical property which renders the cultivation of the gypsum soils so difficult in the Pecos Valley. The New Mexico Experiment Station has shown that, in itself, the gypsum is not harmful to plant growth, but in districts where water contains alkali salts the cultivation of the gypsum can not be recom- mended. One of the most serious difficulties in the way of reclamation of gypsum lands would be their proper drainage. Canals, ditches, and laterals when passing through the gypsum are found to lose quantities of water by seepage, both lateral and downward. Cavities and under- ground channels are dissolved out in the gypsum, which makes it very Report No, 64, U. S. Dept. Agr. Plate V. CARLSBAD AREA WATER SUPPLY. 69 difScult to drain the soil. Open drains have been seen in which the water would flow in from one side, directly across the drain, and out into the subsoil on the other side. The great depth from which water is raised by the gypsum would necessitate the placing of the drains deeper than usual. Open drains can not be used, for the gypsum erodes so easily that the drains are not easily controlled. The irrigation lat- erals on the gypsum area, if run on a steep grade, cut deep trenches and fill up the subsoil with the water which seeps from their sides and bottoms. Plate VI shows such a lateral, which has cut a trench in places 10 feet deep in the gypsum. The virgin gypsum carries small quan- tities of alkali salts in nearly all places examined, but at no point was an accumulation of alkali salts encountered. A very different state of affairs is seen after the gypsum has been irrigated. The top of the soil soon becomes crusted over and the growth of agricultural phints pre- vented. Even whei-e not cultivated, but in the vicinity of ditches and irrigated fields, the subirrigation of the gypsum is so great as in a few years to destroy all vegetation which can not withstand quantities of alkali. From our present knowledge of the subject, the cultivation of pure gypsum soils is to be discouraged in the Pecos Valley. The Water Supply of the Cakslsbad Disteict. The irrigation water of the Carlsbad district is obtained by storage of the Pecos water. It has been shown that the normal flow of the river is all from springs situated along its middle and lower courses in New Mexico. This water contains more or less alkali in solution all of the time, and in standing in reservoirs the evaporation from the surface serves further to concentrate the water. The evaporation in Lake McMillan has been found to be as much as 4 J inches per week, and, though authentic records are not at hand, the evaporation from a water surface is estimated to be about 1 feet of water per year. This great evaporation from the surface of a large reservoir is a very im- portant item in the engineering of a storage system for the southwest. Estimates by the engineer of the Southern Canal system at Carlsbad during the summer of 1899 showed that during several weeks of May and .June the evaporation from the large reservoir exceeded the inflow from the Pecos by as much as 25 cubic feet per second. The inflow from the Pecos is at a minimum during the winter and early spring, and the evaporation during this time continues to concentrate the water, so that when the water is first needed for young, tender plants, such as sugar beets, it is in its worst condition. The first floods, which occur generally in May or June, bring down large quantities of salt which have accumulated in the salt drains of the upper Pecos. These floods seldom improve the condition of the water. Later floods fur- nish the purest water of the season, and it is upon these floods of pure water that the farmer should depend, if possible. 70 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. From analyses by Prof. E. M. Skeats, of Carlsbad, the condition of the water has been found to vary from 510 parts soluble matter in 100,000 parts of water found in the first floods, to 243 parts found in the later floods, with an average of 310 parts. Professor Skeats gives the following as the average composition of the soluble matter in the Pecos water : Parts per 100,000. NaCl 98 CaCOs 11 CaSO, 130 MgSO^ 49 KjSO, 5 Silica, alumina, and iron 2 Water of crystallization, etc 15 Total salts 310 Of this it will be seen that 152 parts per 100,000 are salts likely to accumulate in the soil solutions and injure plants. The application of 2^ acre-feet of water would add 10,000 pounds of soluble salts to the soil, and, since 15,000 pounds can be taken as the maximum amount allowable in the surface foot of the soil, IJ years only would be required so to fill the soil with alkali that agriculture would be uuprof itable. This salt, however, is not all deposited in the surface foot, but is distributed in the soil as deep as the water penetrates, and if enough water is added to wet down to the level of standing water i^art of the salt left from the evaporation of the previous irrigation is washed away. During June, 1899, a sample of the water from the main canal near Carlsbad was taken. A chemical analysis of this sample was made by Dr. Prank K. Cameron, chemist of this division, with the following result : Parts per 100,000. CaSOi 161.64 MgSO^ 2.16 Na^SO^ -> 59.98 NaCl 168.16 391. 94 Undissolved after evaporation 43. 28 Total salts 435.22 The water at the time this sample was taken was said to be so con- centrated as to injure young sugar-beet plants. The osmotic pressure of a solution is the determining factor in its action toward plants ; the water at this time contained about one-half as much salt in actual solution as could be endured by most plants. To this salt was added that already within the soil, so that the delicate rootlets were immersed Report No. 64, U. S. Dept. Agr. Plate VI. t >«^ " CARLSBAD AREA — WATER SUPPLY. 71 ROSWELL Ho^ CHAi^^S_C£U»TrJ _ z ^^ ^. fOO)' COUNTY Spring- Lake \Lake McJdilltta ,. _ ^fl^SRi^- . NEW JyLEXicp 'tt. be Surprise fi au-rta ahtne parta of aotttbt-e natter tn JOO.0OO parts ot water ■Irtaa -HAQlRMflN CANAL TEXAS s PECOS CITV5 104. LONBITUDE WEST Or GREENWICH. Fio. 5.— Sketch map of Pecos Valley, showing increase in salt content of water aa it flowa down the Talley. (The figures show the salt content in parts per 100,000.) 72 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. in a solution almost as concentrated as they were able to stand. Under tbe intense beat of tlie sun of soutbern iSTew Mexico this solution soon evaporates so much near the surface that it concentrates beyond tbe limit of plant growth. In the solution outside tbe rootlets tbe osmotic pressure becomes higher than the osmotic pressure inside the rootlets, and the plant dies of thirst though in tbe i)reseuce of plenty of moisture. Figure 5 gives a graphic idea of tbe salt content of the water in different parts of the valley. Underground Water. One of the maps accompanying this report shows the deptli to water, measured from the surface of the ground. The map indicates by the shades the three conditions of tbe land: («) Water more than 10 feet below tbe surface; (It) water between 3 and 10 feet; (c) water 3 feet or less. The general condition of the vegetation throughout tbe area watered by the Southern Canal system is not healthy. Crops are more subject to disease than is the case under similar conditions in humid climates. Sugar beets are subject to a root-rot; fruit trees suffer from something similar, and even alfalfa has a disease whicli attacks the roots. This unhealthy condition can be attributed to the condition of the irri- gation water. Plants with their roots immersed in a solution of the concentration of the irrigating water from the Southern Canal are weakened and the germs of disease, which seem to be present in all of the soil, are enabled to attack the plant, weakening it further. This root disease is difl&cult to deal with, since the application of sprays to the roots is an impossibility. Some benefit has been claimed from the application of copperas around the roots of the trees. The depth of wells over the greater part of the district was at least 40 feet originally; at present water can be found over nearly the whole of the irrigated land at less than 20 feet. In some of the wells which were dug 40 feet eight years ago water stands now at a depth of 5 to 10 feet during the irrigating season. In the winter, after the water has been taken out of tbe canals and laterals, the water slowly lowers in tbe wells. As soon as the water is turned in the main canal tbe level in the wells begins to rise, even before tbe irrigation com- mences on the lauds around the wells. That tbe loss from canals and laterals is the greatest source of seepage water in irrigation districts is a fact which is coming to be generally recognized. There is without doubt much seepage water from overirrigated land, but tbe loss from constantly running canals and laterals far exceeds this. The loss from the canals of tbe Carlsbad district, while not great when compared with the loss of canals of other districts, is great enough to cause serious damage. The canal in crossing the beds of gravel and conglomerate loses aijpreciably by seepage. The following measurements of loss in Report No, 64, U. S. Dopt. Agr. Plate VII. CARLSBAD AREA — UNDERGROUND WATER. 73 conglomerate will illustrate this point. In running tlirough the con- glomerate directly west of Carlsbad the canal loses 4 cubic feet of water per second in a distance of 1 mile. While this loss is not great when compared with some Colorado canals, as reported by Carpenter, yet the effect of this alkaline water has been seriously felt in the flat to the west of the town, immediately along the foot of the gravel hill. The underground- water maiJ shows au area in which water approaches to within 3 feet of the surface. The greater part of this water is the seepage from the canal, while part is due to the laterals in the town. The water has accumulated to such a degree and has become concen- trated to such an extent by evaporation that the park trees and cotton- woods have suffered and many of them have died. A deep drain around the town would largely prevent this damage. Along the foot of the gravel bluffs, in sees. 8, 9, 10, 14, and 15, T. 22, E. 27 E., is another area in which water has seeped out from the gravel bluff, swamping the land. Jhis water flows under the higher lands immediately south, and no doubt has its origin in the canal. Some of the seepage water comes from the irrigation on this upland, but this source seems inconsiderable when compared with the loss from the canal. A third area of wet land on sec. 25, T. 22 S., E. 27 E. can also be attributed to seepage from the conglomerate. On the edge of this flat water originally stood 40 feet below the surface; while at present, dur- ing i^art of the irrigation season, it is unsafe to attempt to drive a horse along the roads through the flat. The other areas, shown with water near the surface, are either in the area of gyijsuni soil or are immediately adjacent to the gypsum. The loss in canals and laterals in the gypsum is great. The measure- ments in the following table illustrate such loss: Measurements of loss in canals and laterals by seepage in gypsum. Lateral from gate 32 loses 1.64 cubic feet per secoud in IJ miles. Lateral along south aide sec. 12 loses 1.71 cubic feet per second iu 1 mile. Lateral from head gate 18 loses 0.89 cubic feet per second in i mile. Main canal between gates 26 and 27 loses 2.62 cubic feet per secoud in IJ miles. All around the canal for a short distance below it the water stands near the surface, coming by direct seepage from the main canal. Along sections 31, 32, and 33 the canal loses 2.6 cubic feet of water iu IJ miles. This water has seeped out, forming ponds below the canal. A number of photographs illustrating the condition of the country and showing the ponds formed from seepage were secured. One of the photographs shows a stream flowing approximately 2 cubic feet per second, which seeps from the canal. Attempts have been made to stop the seepage from this piece of canal, but so far with- out success. The canal company contemplates abandoning this piece of canal and running the water directly into Black Eiver, from which it will be taken by a canal to supply the lands south of Black Eiver. 74 A aOIL SUEVEY IN THE PECOS VALLEY, NEW MEXICO. Alkali in Soils. The alkali map accompanyiug this report shows three conditions of the soil: (a) Soil with less than 0.25 per cent of salt; (&) soil with from 0.25 to 0.50 per cent of salt; and (c) soil with more than 0.50 per cent of salt. The first represents land which is good; the second, land which contains sufficient alkali to damage crops, but not enough to prevent plant growth; and the third, land which already contains too much salt for crops. The alkali salts are entirely of the white kind — that is, containing no sodium carbonate. The following analyses, by Dr. Frank K. Cameron, were made on samples of crust collected by the members of the Division: Num- ber. Locality. CaSOj. MgSOa. Na.SOi. CaClj. MgCl,. NaCl. 4045 20.56 3.62 3.20 11.29 5.01 11.68 12. 28 7.06 16.53 23.45 6.33 13.39 48.44 27.42 13.56 39.92 37.19 25.76 63.13 54.94 21.48 24. 52 48.31 29.14 18.50 31 35.73 34.35 6.67 26.02 4.60 27.47 32.47 36 51 4040 20.33 4047 26 12 4048 West side Pecos, one-half mile south of Six- 56 61 4052 9.14 4054 Pecos bank, sec. 10, T. 22 S 28 75 4055 Southwest corner orchard, sec. 10, T. 22 S Orchard, sec. 14, T. 22 S 36 01 4056 38. 14 Alkali Hat sec. 14, T.23 S 7.16 28 20 4053 28.26 33.06 4057 42 11 4058 Sec. 14, T. 22 S 40.68 18.20 4061 Sec. 12, T. 23 S 7.52 20.75 26.46 26,95 19.57 4063 24 87 The soils were shown to contain originally only small quantities of alkali salts in their natural state, but at present there are areas con- taining great quantities of such salts. The presence of this alkali may, in nearly all cases, be attributed mainly to the salt which is contained in the irrigation water. The chemical composition compares nearly with the chemical composition of the salts contained in the waters. A comparison of the two maps, alkali map and underground water map, shows the relation between the seepage water and accumulations of alkali salts. In the area of gypsum soil the water is frequently below 6 feet, yet the capillarity is sufficient to bring the salt-laden water to the surface for evaporation. Problems of the Area. This area of land, on which immense amounts of work and money have been spent in reclamation, presents a number of problems for serious consideration. There are two important problems which are in a meas- ure peculiar to the Carlsbad district. These are the methods of irri- gating with an alkali water and the cultivation of gypsum land. The character of the water is the most serious difficulty in the way of prof- itable irrigation. To develop a new supply of water would be an engineering problem diflBcnlt of solution. The use of the present supply is attended with possible loss of crops, especially where the most favorable conditions do not exist. The greater part of the area of Pecos sandy loam has good drainage at present, and the difficulties Report No, 64, U. S. Dept. Agr. Plate VIM. S- o Report No. 64, U, S. Dept. Agr. Plate IX. BARSTOW AREA, TEXAS. 75 encouutered in these soils are at a minimum. Wherever the level of standing water is below 10 feet, there is no present need of drainage; but where (during its yearly fluctuations) the level of standing water approaches the surface of the ground as close as 3 feet, drainage must be installed. Moreover, the accumulation of alkali from the summer's irrigation should be removed by flooding and drainage during the late summer and autumn when the irrigation water is at its best. BARSTOW AREA, TEXAS. A brief examination was made of the conditions at Barstow, Texas. Here the water is taken out of the Pecos Eiver by .a. diverting dam without storage. The water at this point contains more salt than the Pecos water at Carlsbad. Receiving, as it does, the drainage and seepage waters from the Carlsbad irrigated lands and being augmented by alkali springs along its banks, the water at Red Bluff, below the lower limit of irrigation in the Carlsbad district, carried in May, 18D9, 320 parts of soluble matter per 100,000. The waters of the Pecos at the Barstow intake carried in June 390 parts of soluble matter per 100,000, and opposite Pecos City, below all irrigation, it carried 525 parts per 100,000. The application of water of this character to the soil, if long con- tinued, is sure to result in an accumulation of salts beyond the endur- ance of agricultural plants, unless good natural drainage is present or provision made for the escape of the salts by means of underground drains. The soils are derived from the ancient lake basin sediments and are heavier in texture than the soils of the upper Pecos Valley. They originally contained greater quantities of alkali salts and, in fact, there are evidences of the accumulation of salts in beds at depths below the surface. The soil of much of this area is underlaid by gypsum. This in its turn introduces new complications which render irrigation farm- ing in this district extremely dififlcult. The most favorable conditions only can be relied upon to give profitable results in the irrigation of this district. The land has been under cultivation for about five years and, owing to excessive leakage from canals, seepage waters have accumulated to such an extent that much of the land has already been abandoned. All of the lands on the Pecos City side have been abandoned, and at present irrigation is confined to the Barstow or eastern side of the river. The most promising field for such a district would be the cultivation of saltbushes and other plants resistant of alkali. Water from the Pecos, when the water is in its best condition, should be flooded over the land to wash away accumulated salts near the surface. This salt would have to be removed by good nnderdrainage, if intensive farming is contemplated. A district lower down on the Pecos is reported to be entirely aban- doned from the excess of alkali which the water and soil contained. 76 A SOIL SURVEY IN THE PECOS VALLEY, NEW MEXICO. SUMMARY. The foregoiag pages have shown that the condition of the water in the Pecos River becomes more saline as one descends the river. This fact is well brought out in the sketch map of the Pecos Valley, fig. 5. On this map the parts of salt per 1 00,000 parts of water are shown by figures. The seriousness of the alkali problem varies iu direct ratio to the salt content of the water. At Eoswell the alkali is within the soil; the irrigation water is good. By washing the alkali from the soil with the pure water the alkali ma.y be removed and the land thoroughly reclaimed. If once the alkali is removed by the soils being drained, there need be no further fear of damage from alkali. The water of the Berendos contains alkali in sufficient quantity to be harmful to lands, unless well drained. The use of this water has been almost entirely restricted to lands lying close to the deep-cut river chai^nels, where natural drainage is good; therefore the damage from this water has been slight. The water of the Northern Canal contains 0.2 per cent of alkali, and care should be exercised in using it. The soils,at Hager- man contain at present alkali, most of which has been derived from the evaporation of the Northern Canal water. The water of the Carlsbad system contains on an average 0.31 per cent of alkali salts. The soils originally contained very little alkali, the greater part of that now present being due in all probability to the concentration of the irrigation water. The salt already accumulated must be removed bj' drainage, and all further accumulation must be prevented by washing out with fresh water during the winter or when the water is plentiful and fresh. The soils of the Pecos Valley are deficient in organic matter and nitrogenous plant food. The growth and plowing under of leguminous crops is recommended, together with fertilizing with stable manure when such is attainable. In the southern part of the valley, where sugar beets form a money crop for the farmer, chemical fertilizers would be of value. Economy in the use of water and the prevention of leakage from canals or laterals will overcome much of the seepage, which at present is the principal cause of the alkali. The authors wish to acknowledge their indebtedness to Mr. W. Ham- ilton, manager of the Eoswell Land and Water Company, for base maps and for material assistance in the prosecution of the investigation in the Eoswell district; also to Mr. Charles A. Bremond, Mr. Luigi Martini, and others for help and injterest shown in the work. At Carlsbad credit is due to Messrs. Tansil, P. A. Tracy, and W. M. Eeed, officers of the Southern Canal system, for maps and assistance, and to Prof. E. M. Skeats for valuable data, including chemical analyses relating to the district; also to A. E. Goetz and others. During the short time spent at Barstow, Tex., Mr. G. E. Briggs, manager of the Pioneer Canal Company, assisted personally in the A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. By FRANK D. GARDNER and JOHN STEWART. PHTSIOGKAPHY. Beginning in July, 1899, four months were spent in a thorough and detailed study of that portion of the Salt Lake Valley lyiug west of the Jordan Eiver, the object being to map the soils with reference to their character and to 'the extent of, and damage from, " alkali " and seepage waters. The soils were classified according to their texture, and the waters examined with reference to their quality for irrigation purposes. The soils were further studied with reference to their " alkali" content and its effect in varying amounts upon the crops and vegetation. The methods of irrigation were looked into with regard to their success or failure and the ultimate outcome in relation to the condition of the soils, especially with reference to the accumulation of alkali and seepage waters. Nearly 700 borings were made in this district, usually to a depth of 6 feet and occasionally to 9, 12, or 15 feet. In two thirds of these borings the per cent of salt at saturation in each foot section was determined by the electrical method, and in a considerable number of the borings the sodium carbonate (true black alkali) was determined volumetrically. In areas where hardpan forms an important feature, it was mapped and its depth and thickness ascertained. A study was made of its influence toward the action of roots and water together with the probable mode of formation. Full notes were made on the character of the natural vegetation and the kinds and condition of the crops. Wells, both surface and artesian, were examined with reference to their water, the depth to standing water, and the nature of the strata through which they were dug. Drainage and seepage waters were tested with regard to the salts they were carrying from the soil. Only such of the data collected as is essential for a clear understand- ing of the conditions is herein given, the results being embodied in the following text and accompanying maps. Salt Lake YaUey comprises about one-half of Salt Lake County, the remainder being occupied by the Wasatch Mountains to the east and 77 78 A SOIL SURVEY IN SALT LAKE V.ALLEY, UTAH. tbe Oquirrli Mountains to the west. The mountains close in on the south in what is known as the Jordan Narrows, thus ])ractically sur- rounding the valley on three sides and leaving an opening on the north where it borders on the Great Salt Lake. It is one of many similar valleys which, lying between mountain ranges more or less parallel, go to make up the lower and more level parts of the Great Interior Basin, a broad area of varied surface naturally divided into a number of drainage districts. The general form of the Great Interior Basin is triangular, with the acute angle to the south where it extends into old Mexico. At its greatest extremes it is 880 miles from north to south and 570 miles from east to west, including an area of 210,000 square miles. It comprises nearly the whole of Nevada, the western half of Utah, small portions of Idaho and Montana, and large areas in Oregon and in eastern and southern California.' The region is characterized by many short and usually parallel moun- tain ranges, extending generally from north to south, between which are smooth valleys whose alluvial slopes or floors are built up of the debris brought down from the mountains. The character of the climate is plainly seen in the hydrography and vegetation. Perennial lakes occur only in association with the larger mountain masses, while the vegetation of the valley is usually sparse. The annual rainfall varies from 2 inches in the south to about 20 inches in the mountains in the north, while the annual evaporation from a free water surface varies from 60 inches in the north to 150 inches in the south. The larger mountains have timber in their recesses, but only conifers attain sucb size and abundance as to be of economic importance. The climate of the whole area may be classed as arid. The largest subdivision of the Great Basin is the Bonneville Basin (fig. 6), containing 54,000 square miles, or a little more than one-fourth of theformer. Slightly more than two-fifthsoftheBonneville Basin was once occupied by the ancient Lake Bonneville, whose ai'ea was 19,750 square miles. This ancient lake apparently reached its greatest extent during the epoch of maximum glaciation, as is shown by the ijresence of a number of glacial morains which descend on tlie sides of the Wasatch Mountains to the well-marked shore line of the lake when at its highest stage. This shore line, known as the Bonneville shore line, forms a striking feature of the mountain side, both to the east and west, and is plainly visible from all points of the valley. (See Plate X.) The great upheavals which made the mountains and valleys of this region evi- dently occurred prior to the age of the lake, although there have been 'The following brief description of the characteristics and history of the Great Interior Basin and of Lake Bonneville is essential to an understanding of the present conditions. Some of the facts and figures relating to the basin and to the time and extent of the ancient lake have been taken from G. K. Gilbert's report on Lake Bonneville, published as a monograph by tho United States Geological Survey in 1890. PHYSIOGRAPHY. IV minor changes in the elevation sincfe then, as shown by faults at the western bases of the Wasatch, Oquirrh, and Aqui mountains and by the variations in the altitude of different parts of the Bonneville shore line. The altitude of this shore line has been ascertained in a number 114 4Z 41' 4Cf 39" 38" Fig. 6. — Sketch map of Bonneville Basin, showing ancient lake and present lakes. of places, and, except in a few measurements at the extreme south, it varies from 900 to 1,070 feet above the present level of Great Salt Lake, with a mean elevation of about 1,000 feet. All evidences point to the fact that the surface of Lake Bonneville was at that time 1,000 feet 80 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. higher than the preseut surface 6f Great Salt Lake and that its areal extent exceeded the present lake by 13.3 times. At its highest stage the lake overflowed the rim of the basin, the water escaiiing into Suake Eiver. A deep channel was eroded, forming what is now known as Eed Eock Pass. The level of the lake was thns lowered 375 feet, at which point erosion ceased and no further escape of water occurred in this manner. Here the water level was maintained for a considerable period of time, as is shown by the formation of a wcll-marke]i beach line which is known as the Provo shore line. After this stage the lake water fell to its present level by evaporation, occasional interruptions being now noticeable by the number of intermediate shore lines more or less distinct. There is strong evidence of very great climatic change from epochs of great precipitation of snow and rain to periods Avhen high temperature and excessive evaporation occurred. It is not at all improbable that the Great Salt Lake may have been dry at some period, although this question has never been settled. The modern Great Salt Lake is but a remnant of its ancient prede- cessor, which at the time of its overflow maybe considered as a fresh- water lake. Since the Provo stage, however, the water escaped only by evaporation through a long period of time during which the evapo- ration exceeded the precipitation, causing as a result a body of water containing in solution 22 per cent of salt. At the Provo stage the water surface of the ancient lake was 625 feet above the present level of the Great Salt Lake as determined by the altitude of the Provo shore line in anumber of places. (SeePlateX.) Themeandepth of Great Salt Lake is now about 13 feet; therefore, the water over its present area at the Provo stage was forty-nine times its present depth. A com- parison of its area at the two periods shows that at the Provo stage it was 7.4 times its present area. Assuming that the mean depth of the whole body of water at this early period was one half the mean depth of water at that time over the present lake, the volume of water at the former time is found to be one hundred and eighty one times the present volume of Great Salt Lake. If it were further assumed that the lake at the Provo stage was fresh and contained, as the Utah Lake now does, about 80 parts of soluble matter in 100,000 parts of water, we would have, by evaporation to the present volume of Great Salt T • ke, a solution containing 14.5 per cent of salt, providing none of the salt was precipitated during the process of evaporation. As a matter of fact the lake now carries about 22 per cent of salt, which leaves about 7 J per cent unaccounted for. We find, however, that the inflowing streams ■ arry sufficient water to equal the volume of the lake water every 2J j ears; and if it is assumed that they carry a percentage of salt equal t', that carried by Utah Lake, it would require only two hundred and fifty years for them to carry the remaining 7| per cent and bring the saltness of the lake to its present stage. No doubt considerable quantities of the less soluble salts — as, for example, carbonate of lime and sulphates of Report No. 64, U. S. Dept. Agr, Plate X. CO o > Q r • ■ A ' ^3!|f wmi CO o !•* % z ^ %amt < # % r m «». !► > r I'^, £ CD PHYSIOGRAPHY. 81 liuie and soda — were deposited, but the period of time since the Provo stage has undoubtedly been much longer than two hundred and fifty years. It seems, therefore, quite simple to account for the present high salt content of the Great Salt Lake. At a time when the lake was jaa^- high enough to cover the lower levels of the Salt Lake Valley, the water must have been sufficiently salty to have left the soil in a very salty condition upon the subsidence of the water. As a matter of fact, large amounts of salt are found in the lower levels, especially in the lower depths of the soil. Within the memory of the present inhabitants the level of this lake has varied fully 1 i feet. In 1850 it was very low, but for several years thereafter it rose slowly. It then began to fall again, reaching a very low stage from 1861 to 1864. From 1864 to 1868 there was a period of excessive rainfall, during which time the lake rose rapidly, reaching such a height in 1868 that fully 50 square miles of what was mapped ,J 1:^ 1 1 1 1 ^ ^ i Co 1 1 _ _ Z' ■v _ _ / _ _ _ _ , - — — ,» / _ 1-0 — V < s ~ _ _ _ ^ f / s,, _ _^ _ _ _ k f' / \ V- •' *~ - - ^ bi ^ - — / \ _ _ t-/A / /' \ \ .1 ', ^ ■^ — ^ _ o — __ _ _ _ f? FlQ. 7, — Diagram showing mean monthly fluctuations in water level of Great Salt Lake and the rainfall during same period. (The solid line represents level of lake; the broken line the rainfall.) this season as dry land was submerged by its waters. Since that time there have been three distinct periods of rise and fall, but the general trend has been downward, until at the i>resent time the level is about where it was in 1850. Besides this, there is an annual tiuctuatiou, dur- ing which the lake reaches its maximum about June 1 and its minimum about December 1. This annual variation, amounting to from 1 to 2 feet, is the result of a low rainfall from June to September, inclusive, accompanied by high temperature and low relative humidity, condi- tions favoring rapid evaporation and of a greater rainfall and less evaporation during the remainder of the year. This is shown in figure 7. The accompanying diagram (flg. 7) shows graphically the mean monthly rainfall in inches as compared with the monthly change in the level of the lake in feet. The maximum rise'of the lake occurs about two months after the close of the rainy period, and it is about the same leugth of time after the rain again begins before the lake commences H. Doc. 399 6 82 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. to rise, which show.s that the maxiuiuin effect of precipitation t)ii the lake occurs about two months after the precipitation has taken place. Figure s shows thfe semiannual variation in the lake level for the past fifty years, with the accompanying annual rainfall. There is a general agreement between precipitation and the lake variation. The surface of Great Salt Lake is 4,170 feet above the sea level, while au additional 50-foot contour line would include the lower and more level portions of the valley, amounting to one-half of it. Above this the land inclines toward the foothills at the rate of from 50 to 100 feet per mile. To the east the Wasatch Mountains rise abruptly, attaining a maximum height of 7,000 feet above the general level of the valley. Their snow-clad summits and numerous springs are the source of a number of perennial streams that flow across the eastern part of the valley and enter the Jordan Eiver. These streams furnish au Fig. 8 Diagram showing mean annual fluctuations in water level in Great Salt Lake and the rainfall during the same period. (The heavy line represents the level of the lake iu reference to an established bench mark; the light line represents the rainfall in inches.) abundant and good water supply for irrigating the eastern portion of the valley. To the west the Oquirrh Mountains rise less abruptly, reaching an elevation of from 4,000 to 5,000 feet above the valley. The watershed is not so extensive as to the east, and there is less snow, so that the few streams flowing from the canyons are lost, by evaporation or seepage, before they reach the Jordan River. The Jordan Eiver is the main channel through which the waters of Utah Lake and its inflowing streams reach Salt Lake. It is on the Jordan Eiver that the western portion of the valley is dependent for irrigation water. The mountains consist chiefly of granite, limestone, sandstone, quartz, porphyry, and feldspar, and it is from these rocks that the soil of the valley is formed. The mountains abound in ores rich in silver, lead, and copper. Manj' mines are in operation and furnish material for several large smelters located in the valley. CLIMATE. 83 The lake is also a source of commercial enterprise, and according to the statistics of 1890 the value of the annual output of the salt har- vested from the evaporating ponds amounted to $250,000. CLIMATE. The climate of Salt Lake Valley is characterized by low annual pre- cix)itation, low relative humidity, moderate wind movement, moderate temperature, and abundant sunshine. It may be classed as arid. Ac- cording to the United States Weather Bureau records for the past 25 years, the mean annual precipitation at Salt Lake City is 16.2 inches, with a minimum of 10.3 inches in 1890 and a maximum of 23.6 inches in 1875. Previous records, shown in the chart on page 82, show a range in annual precipitation of from 10 to 29 inches. Of the mean annual rainfall a total of only 2.9 inches I'alls during the months of June, July, August, and September. These four months of mean minimum precipi- tation are accompanied by mean maximum temperature and low rela- tive humidity — conditions favoring excessive evaporation. Since this is the period in which crops make most of their vegetative growth, it will be seen how important and necessary is irrigation water for farming. In fact very little farming is carried on except under irri- gation. Statistics for 1894 show that 92,5 per cent of the farms in Salt Lake County are irrigated. Wheat is the only crop that is grown without irrigating and it makes most of its vegetative growth prior to June. The average yield of wheat under dry farming is slightly more than one-third of the average yield under irrigation. The accompanying table shows the monthly and yearly precipitation at Salt Lake City for the past 25 years as obtained from the United States Weather Bureau office at that place. Monthly and annual precipitation, Salt Lake City, Utah, from 1874 to 1899. Tear. Jan. Feb. Mar. Apr. May. Jtiiie. July. Aug. Sept. Oct. Nov. Deo. An. nual. 1874..... 1 Hi. 31 0.90 2.84 0.74 2.42 1.63 0.20 1.74 2.16 0.73 14.67 1875 3.05 o.'n 2.81 1.50 2.91 .90 1.01 .25 1.22 1.36 5.81 2.03 23.64 1876 1.23 1.52 4 2.09 4.30 .09 .83 .92 .42 3.27 .81 1.80 21.28 1877 .87 .38 2.93 2.14 3.49 .80 .02 .28 .90 2.41 1.02 1.11 16.35 1878 1.07 1.87 3.49 .71 2.54 .67 2.63 3.26 2.50 .10 .35 1.34 1.08 .07 .81 .06 3.15 .01 1.39 1.62 .63 .32 .11 3.08 19 75 1879 13.11 1880 .29 1.02 .43 2.37 1.85 .01 .20 .74 .56 .40 1.17 1.90 10.94 1881 1.24 2.44 .88 2.37 2.65 .28 .21 1.66 .43 2.19 1.44 1.24 16.93 1882 1.50 .42 1.12 3.81 .26 2.24 .30 1.61 .37 2.89 .54 .9^ 15.98 1883 1.47 .72 1.75 2.92 .98 .33 .10 .62 .13 2.24 1.78 1,20 14.24 1884 .71 2.23 3.69 2.89 1.78 .33 .27 .73 1.91 .36 .50 2.12 17.52 1885 1.48 1.56 .64 63.47 2.49 2.67 .58 .90 1.29 .59 3.10 .92 19.69 1886 1.91 1.36 2.60 4.43 .06 1.02 T. .59 1.88 1.98 1.79 1.27 18.89 1887 2.36 1.41 .35 1.87 .73 .37 1.23 .69 .55 .30 .25 1.55 11.66 1888 1..52 1.22 2.18 .99 .34 .98 .24 .63 .51 .80 2 2.21 13.62 1889 .73 .81 1.64 1.52 2.97 .01 .08 .92 .52 3.85 1.04 4.37 18.46 1890 3.07 2.05 1.12 .94 .16 .32 .02 .79 T. 1.44 T. .42 10.33 1891 .74 .76 4.66 1.49 .72 1.08 .47 .46 1.19 1.26 .90 2.19 15.92 1892 1.61 .68 2.21 1.90 1.65 1.21 T. .05 .12 1.58 .72 2.35 14.08 1893 .82 1.64 2.68 2.72 1.68 .04 1.19 .71 1.30 1.02 1.18 2.37 17.35 1894 1.31 .83 1.73 1.67 1.22 1.38 .82 .87 2.87 1.01 .28 1.28 15.27 1895 1.32 .85 .81 .73 2.29 .99 .42 .02 .95 .24 2.44 .89 11.95 1896 1.26 .69 1.99 2.53 3.67 .25 1.35 1.47 .52 .70 3.15 .84 18.42 1897 1.16 3.81 2.20 2 .98 .52 .69 .33 .48 1.91 1.19 1.47 16.74 1898 .58 .38 1.71 1.30 4.19 1.45 .18 1.35 .15 1.57 1.95 1.28 16.09 1899 .84 2.98 2.93 .81 2.59 .96 .42 1.06 T. 12.59 84 A SOIL SURVEY IN SALT 1,AKE VALLEY, UTAH. The annual ])recipitatiou in the mountains is greater than in the val- leys, and it is estimated that it includes (5 feet of snow, which lingers on the mountains the greater part of the summer. This is important in relation to irrigation, because it makes the water supply plentiful throughout the season. The annual evaporation from a free water surface in Salt Lake Valley is estimated at 8 feet. Data from one of the salt companies show the evaporation from their ponds to be about 37 inches from June to Sep- tember, inclusive. It should be borne in mind, however, that this is from a saturated salt solution and that the presence of much salt lowers the vapor tension, and, consequently, the rate of evaporation. The evaporation from a fresh- water service would no doubt have been much greater. The mean annual temperature is 51.2° P., with a mean maximum of 76.6° in July and a minimum of 27.9° in January. The mean tempera- ture for from June to September, inclusive, is 70.5° F. The following table gives the mean monthly and yearly temperature for the past twenty-five years. Mean month ly and mean annual temperature, Salt Lake City, Utah, from 1874 to 1899. 1874. 1875. 1876. 1877 . 1878 . 1879. 1880- 1881. 1882. 1883 . 1884. 1885. 1886. 1887 . 1888. 1889- 1890 . 1891 - 1892- 1893. 1894 . 1895 . 1896 . 1897. 1898 . 1899 . Average Jan. 29.5 30.5 27.2 30.3 28.3 28.2 30.8 23.5 24.4 28.4 27.6 28.8 33.2 22.9 21.4 24.8 28.8 25.8 27.6 29 29.7 34 28.8 20.6 33.8 27.9 Feb. Mar. 34.2 36.1 34.1 37.8 39.6 26.2 38.2 26.9 24.1 30.7 36.8 40.8 34.1 38.3 29.8 33.7 30.6 33.6 28.5 26.6 30 36.8 30.9 35.6 29.6 35 37.2 46.9 47.1 49.2 34 42.2 36.6 47 40.8 45.0 37.7 47.2 40 47.7 39.5 38 43.4 39.3 41.1 40.8 40.2 33.6 36.3 40.6 41.1 Apr. 49.8 48.8 49.4 52.6 46.8 54.2 46.7 45.6 48 53.5 48.4 49.2 54.8 55.2 50.4 49.6 47.4 45.9 48.1 51.2 46.4 49 54 50.6 49.8 May. June. 56.7 66.2 66.4 58.2 54.2 60.2 57.2 56.7 57.9 56.6 62.4 60.8 58.6 58.8 61.3 60 55.4 55.1 61.2 57.9 51.4 63.4 63.9 52.6 57.7 67.1 67.8 C8. 6 65.4 68.5 65.2 66 71.2 67.3 70 69.1 64.8 68.9 68.6 68.4 70.3 64.8 62 65.8 67 63.9 63.5 70 66 67.1 65.1 July. 67 77.8 74 76.6 77.4 76.9 76.8 74.2 76 75.4 76.2 74 76.1 78.1 74.9 76.6 78.4 77.8 73.3 76 74.7 74.6 72.6 74.2 71.9 75.9 76.2 75.6 Aug. 74.6 75 71.6 75.8 78.4 75.4 72.8 74.4 76.9 76.8 72.7 73.7 76.1 73.4 74.8 77.4 72.9 74.4 75.4 73.3 75.4 74.8 73.7 75.2 76.8 Sept. 74.7 68.1 65.8 64.4 60.6 68.5 63.6 60.2 64.8 69.7 69.4 65.1 62.8 65.5 70.6 60.6 66 65 69.6 63 01 63.8 64 62.2 65.6 67.1 64.5 56.4 58.2 55.3 51.6 49 52 52 60.8 48.4 46.3 53 64.9 51.8 51.6 54 64.2 48.6 63.2 51.8 52 53.2 64.3 54.4 60.9 48.2 43.2 42 41 40.6 44 36 29.6 33.8 35.5 39.2 42.2 43.8 31.4 42.8 41.6 39 41 44.2 42.2 39.4 45.6 37.8 37.2 43.2 37.3 Dec. 33.9 36.2 27 32.1 29.8 29.4 34.6 33.6 35 32.5 35.2 33.9 36.1 29.2 35.8 39.6 36.2 28.6 27.6 36.3 31.4 26.4 36.2 27.4 25.2 An- nual. 52.4 51.2 51.4 51.9 53 48.6 51.8 49.2 50.8 51.4 62.3 51.6 52.7 53 52.7 51.6 50.6 51.2 50.2 50.8 50.2 51.5 50.2 49.7 51.2 The mean annual wind movement is at the rate of 5.4 miles per hour, with a maximum of 6.4 miles in May and a minimum of 4.3 miles in November. HISTORY OP IRRIGATION. 85 The following table gives the monthly and annual wind movement at Salt Lake City for the past twenty-five years. As a whole, the climate is both pleasant and healthful. Monthly and annual wind movement in miles at Salt Lake City, Utah, from 1874 to 1899. Year. Jan. Feb. Mar. Apr. May. June. July. 3,421 Aug. Sept. Oct. Nov. Deo. Annual. 1874 4,227 4,394 4,116 4,231 4,407 3,924 4,406 3,240 2,681 3,076 2,272 a 31, 188 1875.... '3,395 '2,' 701 '"i,'492 4,672 3,923 3,983 3,087 3,262 2,031 44, 753 1876.... 2,915 2,460 4,109 3,689 3,537 3,570 4,252 4,910 4,524 3,418 2,887 1,636 41, 967 1877 .... 3,222 2,411 4,219 5,206 6,291 4,531 5,165 4,848 4,783 3,998 3,777 2,978 51, 429 1878.... 2,569 3,568 4,160 5,057 5,229 4,970 5, 635 4,083 3,830 4,406 3,148 2,895 49, 440 1879 3,898 3,336 5,001 4,739 6,410 5,085 4,503 4,783 4,104 3, 694 2,222 3,606 60, 380 1880.... 3,516 3,392 5,003 5,693 5,475 4,634 4,090 3,878 3,001 3,499 2,090 3,801 48, 572 1881.... 3,360 2,072 4,045 3,174 3,773 S,769 4,394 4,199 3,916 4,000 2,205 1,473 40, 370 1882.... 2,100 2,952 3,810 5,009 5,669 4,713 4,321 4,476 3,779 3,895 2,613 2,726 46, 963 1883.... 2,333 2,590 4,136 4,914 3,350 4,168 3,224 3,312 3,793 3,746 3,166 2,984 41, 716 1884.... 3,257 4,075 4,662 4,443 4,613 4,875 4,023 3,492 3,844 3,947 2,422 6,410 49, 063 1885.... 2,410 2,378 2,770 3,006 2,652 3,853 4,366 3,921 3,684 2,823 3,705 2,808 38, 276 1886.... 3,520 2, 832 3,736 4,319 4,534 4,146 4,287 3,210 3,734 4,193 2,730 2,239 43, 470 1887.... 3,862 5,350 3,933 4,877 4,926 4,589 3,800 4,524 3,723 3,414 2,863 3,406 49, 267 1888.... 3,776 3,201 4,736 4,160 4,562 4,989 4,123 4,333 2,947 3,619 2,648 2,441 46, 536 1889 .... 2,256 2,239 4,180 4,758 4,920 4,402 4,808 4,309 3,567 3,430 2,857 4,539 46, 255 1890... 4.095 3,920 4,232 4,196 4,272 4,507 3,602 3,279 2,806 2,745 2,008 2,201 41, 863 1891 .... 2; 217 4, 320 3,300 3,761 4,205 4.135 3,618 4,041 4,860 3,798 3,261 4,368 46,884 1892.... 2,474 2,234 4,935 3,954 3,952 4,601 5,039 4,355 4,313 3,778 5,377 3,612 48, 524 1893 3,695 3,822 4,810 5,741 6,729 5,173 4,532 4,196 4,973 3,876 3,776 3,266 53, 678 1894.... 4,135 4,305 5,273 4,968 5, 387 4,836 4,167 4,358 4,401 4,366 3,187 4,182 63, 505 1895.... 4,571 2,947 5 886 6,638 5,513 4,659 4,464 4,134 5,140 3,966 3,612 3,804 54, 134 1896.... 4,274 3,601 4,382 5,710 4,963 4,460 4,108 4,185 5,002 4,234 4,166 3,239 52, 324 1897.... 2,662 3,742 5,490 5,215 6,020 4,663 4,687 4,238 4,637 4,376 3,557 3, 436 51,723 1898.... 3,169 3,480 5,148 5,018 4,639 4,733 4,334 4,486 4,432 3,878 3,679 3,799 50, 675 1899.... 4,449 3,343 6,044 5,366 5,058 4,664 4,348 4,913 3,748 a 40. 923 81,130 81,231 111,642 121,212 123, 267 117,053 111, 134 108, 793 106, 264 92, 867 78, 193 79, 091 1, 210, 777 AYerage 3,215 3,250 4,462 4,662 4,741 4,502 4,274 4,184 4,049 3,715 3,128 3,164 6 47, 444 a For 9 montlis. b For 24 years. HISTORY OF IRRIGATION. Modern irrigation in the United States began in Salt Lake Valley, Utah, when the Mormons settled there in 1847. One of the first under- takings after reaching the valley was the diverting of the water of what is now known as City Creek and the irrigation of a few acres of land planted with seed bi'ought with them on their long and perilous jour- ney from Illinois. Traces of irrigation antedating the Salt Lake Valley undertaking are to be found in Arizona in the systems (long since abandoned) of an extinct race of aborigines, and in southern California, where irrigation was practiced by the mission priests. An historian of the Mormon Church describes the present site of Salt Lake City as follows : A desolation of centuries, where earth seemed heaven forsaken, where hermit nature — watching, waiting — wept and worshipped God amid eternal solitude. Charles Brough, in his Historical and Political Studies on Irrigation in Utah, says: The transformation of this sterile waste, glistening with beds of salt, and soda, and deadly alkali, seemed impossible. These quotations give an idea of how the conditions appeared to the first settlers. In the year 1848, 5,153 acres were put under irrigation, 86 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. and the amomit of land under irrigation in this locality has since raj)- idiy increased. The growth of Salt Ijake City was very rapid, and iu 1850 the population numbered 11,354. During this year there were over 16,000 acres under cultivation. In 1852 the assessed value of property was $400 per capita. To-day the city possesses a population of 70,000. The broad streets are lined with rows of stately trees and the comforta- ble homes are surrounded by luxuriant lawns. The rural districts are populous; the farms are small and are char- acterized by an intensive and diversified form of agriculture. In the whole State of Utah the average size of the irrigation farms is 27 acres. According to statistics gathered in 1894 by the State statistician, there were 2,195 farms in Salt Lake County, 90 per cent of which were cultivated by the owners. Of the total number of farms only 14 per cent were incumbered by mortgage. Notwithstanding the success that has been attained, serious damage has occurred in places through the accumulation of seepage waters and alkali. Districts once successfully farmed have been abandoned and the attempts at reclamation have failed, because suitable methods have been wanting. Damaged lands are more apparent now than formerly, and demand for methods of preventing such damage and for reclaiming waste land is greater than ever before. The earliest irrigation was principally on the east side of the Jordan River, the irrigation waters being obtained from the numerous small streams issuing from the canyons of the Wasatch Mountains. The canals, always small, were constructed by cooperative labor, coopera- tion being the watchword of the Mormons and even to the present time predominating in all lines of mercantile pursuits. A number of farmers owning land along a stream joined together and by their collective labor constructed a canal that brought water to all of their farms. The distribution of the water was proportional to the amount of land owned by each. The advantage of this method was that it gave water to each farmer without expenditure of money and without waiting. The canals were crudely constructed and no provision was made against leakage. Water was turned into the canals in the spring and not turned out until fall, in some instances even running throughout the year. As a result the large amount of waste and seepage waters did much damage to lands lying below the neighborhood irrigated, and at the present time a large area of land immediately south of Salt Lake City and adjacent to the river is much affected by seepage waters and alkali. On the west side of the Jordan River the earliest attempts at irriga- tion were on the Jordan meadows or river bottom lands, the water supply being obtained from the Jordan Eiver by means of small canals. Subsequently the Brighton and North Point and the North Jordan canals were run upon the first terrace above the river, and following these were the South Jordan and the Utah and Salt Lake canals on the second and third benches, respectively. HISTORY OP jR&I&AtlOiJ. 87 As is frequently the case, the irrigation on the benches caused an accumulation of seepage and alkali on the river bottom land, so that much of it has been abandoned. The largest and most seriously dam- aged area, however, is just south of Twelfth Street road, and comprises a strip of land varying from half a mite to a mile and a half in width, and extending 10 miles west from the river. Here the seepage and surplus waters from the outer extremities of the Utah aud Salt Lake, the South Jordan, and the North Jordan canals have collected to an alarming extent. Indeed, the damage has gone so far that a chain of lakes has formed, presenting a water surface of fully 1,000 acres. The area affected is not less than 10 square miles. That portion of Salt Lake Valley west of the Jordan River which is at present under irrigation includes about -10 square miles, and covers a strip about 2 miles wide, bordering on the river and extending through Ts. 2 and 3 S., R. 1 W., together with another narrower por- tion at the north, which bends to the west through T. 1 S., Es. 1 and 2 W., nearly to the point of the mountains. It consists mainly of ter- races, one above another, and has a slope toward the river or to the north of 50 to 100 feet per mile. In addition to the above-named canals there is the surplus canal, from which the North Point Consolidated Canal is taken. The latter conducts water to the low-level area north of the base line, but is little used for irrigating purposes because of the unsatisfactory results of aiiplying water to this level, salty land. The irrigation canals on this side of the river have an aggregate capacity of about 600 cubic feet per second, but less than half of this amount is required or used on the 25,000 acres under cultivation. On the low-level area, between Salt Lake City and the lake, many attempts have been made and much money expended in the endeavor to success- fully irrigate the laudj but, with a few minor exceptions, the attempts have all proved failures. The canals are owned, for the most part, by the owners of the land under irrigation, and the only paid officer is the "water master," whose duty it is to attend to the equitable distribution of the water to the shareholders. At stated intervals, along the main canals, laterals are taken out to supply the farms along its course. Bach lateral has a head gate, the opening or closing of which is controlled by the water master, aud the size of the opening is varied according to the number of shares supplied by the lateral and the total water supply for the canal. If the water supply is plentiful, the gates usually remain with a certain sized opening throughout the season, nnd the water is per- mitted to flow continuously. Bach shareholder is entitled to use all of the water flowing in the lateral for a stated number of hours and at stated intervals, according to a schedule agreed upon at the beginning of the season. 88 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. SOILS. The area surveyed in 1899 includes all of that part of Salt Lake Valley lying west of the Jordan River, and is equal to about two-thirds of the entire valley. It extends westward to the foothills of the Oquirrh Mountains and the Great Salt Lake and northward to the lake. The area, roughly estimated, is 14 miles east and west by 28 miles north and south at its greatest extremes, and includes over 250 square miles. Toiiographically the area varies in elevation from the present level of the lake, which is 4,170 feet above sea level, to about 4,700 feet at the foothills. A contour taken 50 feet above the lake would include the northern half of the district, which is comparatively level. The por- tion above such a contour inclines toward the mountains at the rate of from 50 to 100 feet per mile. The drainage is into the Jordan River or directly into the lake. The soils have been formed by material brought down from the moun- tain sides and by sediments from the aiicient Lake Bonneville, all of which have been materially modified by inflowing streams from the mountains and by the vacillating shore of the lake. Soils formed in this way are usually heterogeneous, and these soils form no excep- tion to the rule. In the lower part of the valley the sediment is very deep, no rock or gravel being found at a depth of a hundred or more feet. As we get near the foothills, gravel and rock are plentiful and often croj) out at the surface. Here there is little or no lake sediment apparent. The soils are fertile, but in the natural condition support only a meager vegetation, because they are either too dry or too salty. On the higher portions, where there is little salt present, sagebrush forms the chief growth, while in the lower areas, where there is more moisture and much salt, salt-loving perennials, such a greasewood and "mutton sas," abound. The following classification of the soils is based on the judgment of the field experts, typical samples of the various types being sent to the laboratory and analyzed, not as a basis for classification, but in order to obtain an explanation of certain characteristics as they appeared in the field. The chief basis of the classification is texture, as determined by the feeling and appearance of the soil, and it will be seen by studying the analyses of samples of the different types that the judgment of the experts is quite as accurate as the analysis itself. The classification is based chiefly on the characteristics of the first foot in depth of the soil, although the underlying stratum is sometimes considered, as in the case of the Bingham gravelly loam, where the gravel is sometimes absent in the top foot, but occurs in the second or third foot. The soils have been classified under eight types, in the order of the magnitude of their respective areas, as follows: JORDAN SANDY LOAM. 89 1. Jordan sandy loam. 2. Bingham gravelly loam. 3. Jordan loam. 4. Jordan clay and clay loam. 5. Jordan meadows. 6. Jordan sand. 7. Bingham stony loam. 8. Salt Lake sand. JORDAN SANDY LOAM. This loam, shown on the soil map by the orange color, comprises about 30 per cent of the entire district, and is the most important of the various types of the soil, both in extent and quality. It is a light, sandy loam, varying from one to several feet in depth, the texture of which is shown in the accompanying table of mechanical analyses. Mechanical analyses. .g.s £> \a 5 w S in o g No. liOcallty, Description. h li l1 Is a 1 & o s ■3 1 5 i o" 1 °a ■aa o a d la IS 1 a a o d ° d i o d s d 2 CO Ji d i- 3 Jordan sandy loame to IIS inches in depth. Per Per Per Per Per Per Per Per Per Per Per cent. cent. cent. cent. cent. cent. cent. cent. cent. cent. cent. 4303 SW.i 860.17,1.1 N.,T1.1W. Dry level land-. 0.49 L18 1.84 0.16 4.17 11.40 27.87 17.76 23.49 2.64 8.79 43H SE.i sec. 31, T. 2 Low level land. .41 .94 2.17 1.10 3.28 7.80 15. 23'23. 46 33.06 2.80 9.47 N., K. 1 W. 1 4309 C. 860.24, T. 1 S., Dry level land.. 1.69 1.62 9.20 0.00 .12 .61 26. 87123. 55 23.30 3. 59:10. 87 E.2W. 4369 S. C. sec. 17, T. 1 S.,E.1W. ..-•..do 1.50 1.45 3.31 T. .28 .95 8.84 39.31 29.44 3.28 11.74 4310 C. 860. 20, T. IN., E.2W. do L09 1.79 5.25 .13 3.96 9.61 20.13 16.43 26.36 2.59 13.37 4297 N. C. 860. 27, T. 1 N., E. 2 W. do .89 1.55 4.57 .18 3.42 10.01 23.74 15.65 24.38 2.80 13.55 4365 NE. C. sec. 16, T. 2S.,E.1W. Alfalfa, irri- gated. .30 1.95 5.42 T. .77 4.98 3.11 26. 82 35.85 3.87 14.58 4366 S. C. sec. 1, T. 1 S., E.2W. Subsoilsunder Jor- dan sandy loams. Trees, irri- gated. .44 1.27 4.06 .67 1.06 1.43 9.11 29.85 33.08 4.75 15.03 4370 Loam 24 to 36 inches. Under 4369 1.08 1.49 5.61 T. .48 1.63 8.61 27. 12 29. 42 1 2.60 22.43 4371 Sandy loam 36 to 60 inches. do .66 .99 4.24 T. .82 5.90 42.4130.33 2.79 13.10 4367 Loam 12 to 24 Under 4366 .44 1.27 5.16 .35 l.ll' 1.40 5.84 30.05 31.11 5 18.21 inches. 4368 Fine 8and=8andy do .46 .90 3.94 T. .31 .34 3.10 46.38 32.82 2.72 9.54 loam 24 to 48 inches. The analyses of the eight samples of soil in the above table, the samples being taken from the first foot in depth, probably represent the range in texture for this type of soil. The clay content is compara- tively low, ranging from a minimum of 8.79 per cent to a maximum of 15.03 per cent with a mean of about .12.2 per cent, which may be taken as representing the average clay content of this type of soil. By far the larger part of th6 separations occur under the heads of very 90 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. f:no saud and silt, which give an average of 24.1 aad 28.6 per cent, respectively. The soil varies somewhat in underlying strata, but the most usual profile is 2 feet of sandy loam, 1^ feet of loam, and 1 foot of fine sand, underlaid by clay to considerable depth. In the above table Nos. 4370 and 4371 show the texture of the loam and sandy loam (the latter approximating very fine sand) which underlie soil No. 4369, while Nos. 4367 and 4368 represent the loam and fine sand underlying soil Ko. 4366. The clay, which usually occurs at a greater depth than these samples represent, corresponds in texture to the analyses given for the same under Jordan loam on page 95. Subsoils of this texture permit a ready movement of the ground water, and should therefore be easily under- drained by placing lines of tiles 150 feet apart. As a rule the Jordan sandy loam overlies the Jordan loam, on which it generally borders, and it occurs mostly in the irrigated district and the low land to the north. That portion of it lying above the irrigation canals, as well as that in the irrigated district wherever the water table is 10 or more feet below the surface, is free from any excessive amount of salts. On the low area to the north, however, it generally contains considerable salt, especially in the lower depths. The native vegeta- tion on this part of it consists mainly of greasewood and shad scale (perennial bushes growing from L to 4 feet in height) on the drier and less salty portions; and of mutton sas, salt grass, and various small salt-loving annuals on the moist and more salty places. The accompanying table shows the per cent of salt in solution when the soil is saturated with water for various depths and places in the Jordan sandy loam, the percentage being calculated on the water-free soil. Table showing the soluble salt content at saturation for various places and depths in Jordan sandy loam. Numberof boring. Depth, in feet. Depth to stand- ing water. 1. 2. 3. 4. 6. 6. 7. 8. 9. 200 Per ct. 0.06 .06 1.51 .76 .18 .35 .11 .14 .40 .09 Per ct. 0.05 .05 1.17 .63 .15 .54 .68 "'".'is' Per et. 0.05 .06 .74 .40 .19 1.02 .88 1.01 .76 .85 Per ct. 0.06 .05 .50 .35 .25 "i'.ai 1.12 "i.io Per ct. 0.08 .09 .62 .29 .25 1.06 6.35 1.39 1.04 1.54 Per ct. 0.08 .15 .72 .28 .20 Per ct. 0.10 Per ct. 0.11 Per ct. 0. U Feet. 10-t- 20 4 ?* 206 16 99 300a 651 1 6+ 6+ 7 629 1.64 2.00 644 517 626 2.22 6-t- " a Irrigated. b Sand pocket. Borings Nos. 200 and 206 represent about the mean salt content of this type of soil on the good irrigated land. Here the salt rarely exceeds 0,1 per cent above a depth of 5 feet. While there is a gradual increase in the salt as we go down, it never occurs to an alarming Report No. 64, U. S. Dept. Agr. Plate XI. JORDAN SANDY LOAM. 91 extent, although there may be sufficient to wash out into seepage areas below, and thus cause damage by. its accumulation. Borings Kos. 16, 99, and 651 show the condition on the low land where the water table is within i or less feet of the surface. Here there is a comparatively large amount of salt present and Nos. 16 and 99 show an accumulation in the surface foot, which is usually the case when the water is 3 feet or less below the surface. These bor- ings were made during the driest part of the year, and it is probable that the water table is nearer the surface during the time of the year when more rain falls. Boring No. 300 is in a favorable location on the low laTid where the water, before irrigation was undertaken, was nearly 10 feet below the surface. It has been irrigated for three years and is now planted in trees. The salt content is fairly uniform in its distribution and, while not present in great quantity, there is sufficient to cause serious trouble if allowed to accumulate in the surface portion of the soil. This it is likely to do if the water table rises much above its present level. Borings No. 629, 644, 517, and 626 are representative of the amount and vertical distribution of the salts for the unirrigated low-land part of this type of soil, where the depth to standing water is 6 or more feet, which is usually the case. In these samples the aggregate amount of salt to the depth of 6 feet is considerable and will correspond to the area shown on the map by an appropriate color, where the salt content to a depth of 5 feet ranges from 0.6 to 1 per cent. A small portion of this type of soil, however, falls within another area on the map, where the salt content exceeds 1 percent, while, on the other hand, some of it falls in the blue area, with less than 0.6 of 1 per cent. The first foot usually carries a relatively small amount of salt and indicates that crops could be successfully grown. As we go deeper, however, the amount of salts increases very rapidly and the second foot is usually about the limit for alfalfa, while the third foot almost invariably contains too much to permit the growth of any agricultural crops. At 6 feet there is usually about 2 per cent of salt. This distribution of the salts probably occurs because the rains wash them downward more rapidly than they are returned toward the sur- face by evaporation. If irrigation water were applied, the salts would continue to move downward, provided the water table remained at its present depth. Unfortunately, however, the application of M'ater inva- riably causes a rise of the water table, and if the application be con- tinued over a considerable area the water table comes sufficiently near the surface to cause excessive surface evaporation, which results in an upward movement of the salts and their consequent accumulation at the surface. The present surface conditions as regards the amount of salt are fairly good over considerable areas, but in order to improve or even maintain these conditions under irrigation thorough underdrainage is imperative. 92 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. The Jordan sandy loam is easily cultivated and is sufficiently fertile to produce almost any class of crops. It forms the most valuable por- tion of the low salty area, because of the ease with which it may be reclaimed by underdrainage and washing. Owing to the light texture of the subsoil, the lines of drains could be farther apart than in the heavier soils, and it would therefore be less expensive to drain. The diagram on page 93 (iig. 9) shows sections in various directions in the valley and illustrates the constitution of the soils to a depth of 6 feet. BINGHAM GRAVELLY LOAM. This type of soil is next in extent to the Jordan sandy loam, and comprises 60 square miles, or L*4 per cent, of the entire district. Except- ing about 2,000 acres immediately under the Utah and Salt Lake Canal, principally in the southern part of T. 1 S., R. 2 W., it all lies above the irrigation canals and is too elevated to be irrigated by any of the pres- ent water supply. There has been a scheme proposed for raising part of the water passing through the Jordan Narrows by hydraulic means and constructing another canal above and parallel with the Utah and Salt Lake Canal. The height to which it would be profitable to raise water for irrigation purposes by such means would take in only a nar- row strip of this upland, owing to the steepness of the slope toward the mountains. At Herri man there are about ] 00 acres irrigated by water taken from Butterfield Creek, and in the northwestern part T. 2 S., E. 2 W., there are a few small springs which serve to irrigate a very limited area. There are also a few small farms along Bingham Creek irrigated by its waters. During freshets considerable water comes down from the canyons of the Oquirrh Mountains, and, in a few instances, some of it is diverted and used for irrigation. This, however, is very unsatisfactory, because at times when water is most needed no water is to be had. The only pos- sible means of successfully irrigating any considerable area of this land is by storing the water from the mountain streams in reservoirs con- structed for that purpose. The Bingham gravelly loam comprises the area shown on the soil map by the brown color. It is always underlaid by gravel at within 3 feet of the surface, and the small gravel generally papears at the surface in greater or less quantities. The following table gives the mechanical analyses of four samples of this soil to 12 and 15 inches in depth. Report No. 64, U. S. Dept t PLATE XII. ^^ ■■¥:;■ 3. o :; r =• o = s ■St . "•?.' ' ^^ p^f^^l^ Vt^:,^i^^5t^^' « . » 4a SOIL SECTIONS. 93 1 f- ! ,•<■■■• » o 5! ir-F- 94 A SOIL SURVEY IN SALT LAKE A'ALLEY, UTAH. Mechanical analyaea of Bingham gravelly loam soils. [j.^ tn o ^ 1 o lO ^ a d c in d to 0. oTt s a § o g d "Sa 1 ^ ■2 S =1 . c a o ■M o No. Locality. Description .3 P. 1 a la o s c ga o jl 1 1 gs H m ® Ed B g 5° '5 d "« d 3 <0 p ^ a o h o 1 o .3 t o a .^ '■D a 3 Per Per Per Per Per Per Per Per Per Per cent. cent. cent. Cflnt. cent. cent. cent. cent. cent. ceni. 4362 SAT. J sec. 16, T. 3 S., E. 2 W. Eolling land to 15 inches, 25.5 per cent coarse gravel. Rolling land to 15 2.25 3.96 1.50 2.71 5.81 8.87 27.93 28.08 5.13 13.53 4363 S. C. 860. 24, T. 2 2.90 4.16 0.66 1.77 7.71 8.50 28. 71 22. 19 4.38 16.98 S.,E. 2W. inches, 9 per cent coarse gravel. 4312 SW. i sec. 2, T. 2 S., E. 2 W. Rolling laud to 12 2.39 4.53 7.57 2.64 2.88 7.29 13.06,37.12 4.34 18.15 inches, 64.5porcont coarse gravel. 4313 SB. i sec. 34, T. 3 Eolling land to 15 3.04 5.06 .33 .77 2.48 4.09 16.73 38.91 5.08 22.06 S., E. 2 W. inches, 4.4 per cent coarse gravel. The analyses show the texture of the flne earth after all gravel larger than two millimeters in diameter has been taken out. The clay ranges from 13.5 to 22.6 per cent of the fine earth. The coarse gTa\ el varies from 4.4 to 54.5 per cent. The gravel is small and more or less rounded and interferes little with the cultivation of the soil. Belo^y 3 feet, and sometimes at even a less depth, the gravel becomes large and occasionally gives place to bowlders and rock. The most usual i^rofile to 6 feet in depth is 18 inches of gravelly loam, underlaid by large gravel. This tyjie of soil is usually free from noticeable amounts of salt, as will be seen by referring to the salt map. The native vegeta- tion consists largely of sagebrush, with some rabbit bush, grass, tum- ble weeds, etc. A considerable percentage of this type of soil is dry farmed to wheat. Asa rule, the yield is small, but in years of abundant rainfall it some- times exceeds 20 bushels per acre. The land slopes rapidly toward the mountains and has many deep washouts, which seem much larger than would be required for the natural escape of the drainage waters. These were probably formed by cloudbursts, which at some period visit most of the areas iu the Bonneville Basin. JORDAN LOAM. This type of soil, while ranking third in extent, is perhaps second in importance, as it mostly lies within reach of the present irrigation water supply. It comprises about 50 square miles, or 20 per cent of the total area, and four-fifths of it lies below the present canal sys- tems. Of this portion, however, there are numerous isolated areas, occurring in the area of clay flats near the shore of the lake, which, owing to their irregular forms, small size, and location would be rela- tively expensive to irrigate. The main body of this type of soil, liow- JORDAN LOAM. 95 ever, could be easily irrigated and, as demonstrated by that portion now under cultivatiou in the irrigation district, would prove excellent land if put in proper coudition. This type of soil is shown on the soil map by the areas in solid red. It varies much in depth and uuderlying strata; the most usual profile, however, is 3 feet of loam, underlaid by clay which contains frequently pockets or strata of sand. The accompanying table of mechanical analyses shows the texture of the surface foot in six localities and the character of the subsoil under- lying two of them. Mechanical analyses of Jordan loams. n • >» Ift o i-i o m ^ .M (C u c< +3 +i a B d o o 1 '3 • S 1 i o o in d o d fa 6 o No. Locality. Description. a p. ss 1 s IN f 10 « ia aim = a •5° o i d Is 1 O 1 O 1 Per Per 1 Per a S Per 5 Per Per Per Per Per Per Per cent. cent. cent. cent. cent. eent. cent. cent. cent. cent. cent. 4331 NW. C. sec. 34, T. 1 N., E. 2 W. Dry level land.. 0.80 1.49 3.72 0.31 0.61 2.51 7. 03 26. 95 37.42 4. 14.70 4318 S. C. s'ec.'i?, f. 3 S., E. 1 W. Good alfalfa land irrig. .56 2.30 5.05 .12 .16 .38 2. 62 20. 92 1 37.73 6.73 22.68 4324 S. C. sec. 18, T. 1 S., E. 2W. Low, wet by springs. 2.38 2.39 14.13 .10 .47 1.08 2.6812.45 28.59 12.98 23.16 4325 NE. Jsec.35, T.l N., E.2W. 2f.C.8ec.4,T.lS., E.2 W. NW. i sec. 8, T. 1 N.,E.2W. Dry level land-. .81 2.35 4.89 T. .17 .64 4.66 20.68 36.23 6.99 23.51 4364 do 1.08 2.10 6.81 1.57 2,42 3.23 5.67 14.22 29.96 6.49 27.66 4336 Low uneven 1.76 1.07, 8.57 T. .56 1.25 7.45 33.70 12.46 1.06 31. 80 land. SnbsoiU under Jor- dan loams. 4332 I/O am 12 to 24 inches. Under 4331 1.16 1.37 9.41 T. 1.47 2,10 6.53 24.81 30.58 2.38 21.04 4333 Very fine sand 24 to 36 inches. Sand and clay 36 do .95 1.03 6.83 .11 ,66 7.78 42.67 27.04 1.85 11.91 4334 do 1.25 .93 5.64 3.78 2.91 9.33 17.56 17.97 24.18 2.40 13.63 to 48 inches. 1 4335 Clay loam 48 to 72 do 1.59 2.18 8.16 1. 26! . 88 1.16 2.40 13.38 35.49 4.52 28.37 inches. 4372 Clay loam 12 to 24 inches. Clay loam 24 to 36 inches. Under 4364 1.74 2.64 9.22 .49 1.02 2.65 5.08 11.87 24.63 7.10 33.47 4372b do 2.24 2. 15 11. 21 3. 24 2.52 2.49 2.34 5.88 23.83 8.64 33. 14 4373 Clay 30 to 48 inches. do 2.24 2.10 10.23 T. .41 ,26 .41 7.33 24.46 9.23 44.38 The clay content varies from a minimum of 14.7 per cent to a maxi- mum of 31.8 per cent. The latter amount is rather high to be classed as a loam, 30 per cent being usually taken as the upper limit for this class. This sample, however, was easily classed as a loam by the observer in the field, and the fact that it contains a lower percentage of silt and fine silt than any of the other samples analyzed for either this or the two preceding types of soil accounts for the apparent conflict between the analysis and the field judgment. The silt content, being unusually low, counteracts the effect of part of the clay, and therefore gives the sample the characteristics of loam. The mean clay content of 23.9 per cent for these six samples may be taken as representative of 96 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. this type of soil. The other large separations are under the heads of very fine sand and silt, which give -!1.5 and 30.4 per cent respectively. The first four separations show a very small percentage in any of the samples. Samples Nos. 4332-35, inclusive, show the texture of the subsoil to 6 feet in depth under soil No. 4331. The second foot is also a loam, but heavier than the first, which is on the border line between Jordan sandy loam and Jordan loam; next comes very fine sand, which continues about two-thirds of the way through the fourth foot, and below this is clay loam. This character of subsoil, while common under Jordan loam, is perhaps more characteristic of the first type of soil. Samples Nos. 4372 and 4373 show the texture of the subsoil under soil No. 4364. Here the second and third feet are clay loam and the fourth foot clay. This is more characteristic of this type of soil than is the former subsoil, excepting that the overlying loam is of less depth than the average. Subsoils of this character are so heavy that ordinarily they would be rather expensive to drain, on account of the short intervals at which lines of tiles would need to be laid in order to prove effective. The field observations here, however, show that the clay or clay loam subsoil is most usually inlaid by strata of fine sand varying from a fiaction of an inch to a foot or more in thickness. While these strata are continuous for only short distances, yet they occur at such frequent intervals that they would undoubtedly be of material assistance if under- drainage were undertaken. Soil No. 4324 occurs in an area wet by large springs and producing a luxuriant growth of salt grass, which accounts for the high i)ercentage of organic matter. The apparently high percentage of organic mat- ter in the heavier samples is probably in part water of crystallization, which is only driven off by temperatures higher than are required for moisture determinations. On the lowland the Jordan loam lies slightly lower than the sandy loam. It has the water-table rather near the surface and carries a higher percentage of salt. The accompanying table gives the percentage of salt in each foot to a depth of 6 feet for various places in this type of soil. Table showing the per cent of salts at saturation at various places and depths in Jordan loam. [Percentage calculated on water-free soil.] No. of boring. Depth in feet. Depth to standing water. 1. 2. 3. 4. 5. 6, 8. Per cent. 0.07 .05 .58 .49 .68 2.07 .11 .38 .74 1.43 Per cent. 0.07 .05 .23 .21 .41 1.44 .09 .96 1.38 2.37 Per cent. 0.08 .06 .19 .18 .24 1.13 .23 1.17 1.65 2.60 Per cent. 0.09 .07 .19 .16 .16 .86 .54 1.38 2.09 3.46 Per cent. 0.10 .07 .14 .16 .16 .82 .88 2.04 2.09 3.66 Per cent. 0.07 Per cent. Feet. 10-1- 149 447 .18 .17 .16 .92 81 2.36 2.21 3.37 1 3 127 3 ? 623 624 0.67 ? JORDAN CLAY AND CLAY LOAM. 97 Borings Nos. 275 and 149 are typical of the favorable conditions for this type of soil in the irrigated district. Nos. 447, 138, and 127 are representative of the conditions in the southern part of T. 1 S., R. 1 and 2 W., where seepage water has caused an abandonment of the land for farming. Here the water table lies within 3 or less than 3 feet of the surface, and while the total salt content to a depth of 6 feet is only about 0.25 per cent, yet in the surface foot it has accumulated to such an extent as to be fatal to some crops. By removing the water the soil would soon return to a productive state. Boring No. 375 is in a low level area which is watered by large springs. At the time this determination was made the water had been turned off, and the water table was 4J feet below the surface. Ordi- narily it is much nearer the surface than this, and, in fact, the surface is frequently covered by water. The salt content is high and shows an accumulation at the surface, this being always a result of wetness. Nos. 623, 624, 343, and 649 show the range in salt content and its usual distribution on the lowland part of this type of soil when the water table is more than 4 feet below the surface. They all show the mini- mum amount in the surface foot, and a gradual increase as the depth increases. No. 623 shows an unusually small amount, 649 an excessive amount, and 624 and 643 normal amounts for this type of soil. A comparison of the soil and salt maps shows that most of this type of soil which is on the lowland falls within the slate colored area on the salt map — i. e., it contains from 1 to 3 per cent of salt to a depth of 5 feet. Excepting about 2 square miles near the mouth of the Jordan Kiver, none of it falls into the class containing more than 3 per cent; but on the other hand a small per cent falls in the class containing 0.6 to 1 per cent. Like the Jordan sandy loam, this type of soil also frequently shows a comparatively small amount of salt in the surface foot, but as we descend it increases rapidly, and when the soil is put under irrigation the salts soon accumulate at the surface, unless by some means the water table is kept down and the surface evaporation reduced to a mini- mum by the best cultivation or by shading crops. The larger percentage of this type of soil is capable of reclamation by underdrainage and washing, but at a somewhat greater outlay than would be required for the Jordan sandy loam. JORDAN CLAY AND CLAY LOAM. The next type of soil in order of extent is the Jordan clay and clay loam, which comprises about 35 square miles, or 14 per cent of the whole area. Excepting about 1,500 acres, one-half of which lies west of Williams Lake and the remainder southwest of Decker Lake, this type of soil lies from 4 to 8 feet lower than the land immediately adjoin- ing. It is level and wet, and rarely contains any vegetation. It forms what was formerly the iioors of lagoons near the shore of the lake and H. Doc. 399 7 98 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. oftea extends far back into the higher and better laud. In other places it assumes the form of draws extending like wide irregular canals back into the laud for miles. Its distribution is shown on the soil map by the blue color. The accompanying table gives the mechanical analyses of two samples of soil and the subsoil under one of them to 6 feet in depth. Mechanical analyses of clay and clay loam soils. So. Locality. Description. & U .si U ro .2 o •A 1 a t o 1" d $. r-l is i o O d $■ = a' r 00 a s d is 1 o d 2 lO O d to § O d -S CO 1 r.1 g s. . d B Per Per Per Per Per Per Per Per Per Per Per cent. cent. cent. cent. cent. cent. cent. cent. cent. cent. cent. 4345 C. S. 18, T. 1 N. 2W. K. Salt flat 8.89 11.5) 3.94 4.91 2,92 2.28 14. 10 19. 15 4.23 2."). 67 4351 S.30,T.l N., W. B.l Low level land . 2.04 2.63 6.39 .77 .90 2.03 4.37 20.68 27.46 4.35 28.18 (Subsoil underiS46.) 4346 Depth 12 to inches. 24 Clay loam 11.47 2.10 13.65 7.19 6.17 4.94 3.62 7.28 22.78 4.47 16.86 4347 Depth 24 to inches. 36 Yellow clay 7.82 11.02 12.96 4.14 1.88 2.66 1.16 6.47 23.53 3.45 26.87 4348 Depth 36 to inches. 48 Bed clay 5.92 8.24 8.43 2.68 1.06 .64 .64 8.41 27.19 3.74 35.23 4349 Depth 48 to inches. 60 do 6.53 10.20 9.50 8.58 .41 .23 .28 3.27 24.02 3.38 33.50 4360 Depth 60 to inches. 72 do 5.02 5.32 .61 .29 .32 .50 14.42 30.04 10.80 33.74 Sample No. 4351 shows the texture of the small upland area just west of Williams Lake. It is very similar to the Jordan loam, which lies adjacent, but is slightly lower and somewhat heavier in texture. Its salt content to a depth of 6 feet is shown under boring No. 162 in the table giving salt content. The total salt content does not differ mate- rially from that in the adjacent Jordan loam, but it shows an accumu- lation in the upper portions, which is not the case with the latter, where the water table lies at 6 feet, as it does here. The probabilities are that the water table is much higher under this soil during a portion of the season. No. 4345 shows the texture of the clay loam of the flats, 20 per cent of this sample being under the heads of organic matter, and salt. Therefore the clay content should be increased from 25.67 to over 30 per cent, in order to show the real per cent of clay present in the soil alone. The subsoil beneath this is shown in Nos. 4346-50. JORDAN CLAY AND CLAY LOAM. 99 The accompanying table shows the per cent of salt for this type of soil : Salt content at saturation for various places and depths in Jordan clays and clay loams. [Percentage calculated on water-free soil.] Number of boring. Depth in feet. Depth to stand- ing water. 1. 2. 3. 4. 5. 6. 8. 10. 12. 15. 132 Per ct. o.u 1.18 5.58 4.50 10,60 Per ct. 0.11 1.23 4.95 5.06 9.80 Per ct. 0.12 1.12 6.30 6.03 11.40 Per ct. 0.13 .79 6.41 "ib.lb' Per ct. 0.12 .68 5. Per ct. 0.12 Per ct. Per et. Per ct. Per ct. Feet. 2 162 6 355 a 4.44 2.96 1.86 .88 a 645 7.82 9.80 \' 646 9.30 Boring No.. 132 gives the salt for each foot to 6 feet in depth for the area southwest of Becker Lake. The water table at the time of making this boring was only 2 feet below the surface, but the deter- mination shows only a slight tendency to an accumulation of salt at the surface. In other portions of this area, however, the accumulation of the salt at the surface has caused the land to be abandoned. Nos. 355, 645, and 646 give the salt content for three places in the barren clay loam and clay flats, and the amount present is simply astonishing. The above determinations show a range of from 4.5 to 11.4 per cent of salt in solution at saturation in the uppef 6 feet of soil, and it is probable that at such concentration some of the salts may remain undissolved and are therefore not shown by the electrical method, which was used in their determination. Below 6 feet in depth there is a gradual diminution in the salt content, and in boring 'So. 355 it is only 0.88 per cent at 15 feet in depth. Boring No. 646 shows a very high salt content, the average to 6 feet in depth being 10.2 per cent. Allowing 70 pounds as the weight of a cubic foot of this soil, the amount of salt present, to a depth of 6 feet, in one square mile would amount to approximately 1,200,000,000 pounds. To the average mind such large numbers give no adequate idea of the real amount. By reduction we find that the 1,200,000,000 pounds equal 600,000 tons, which, at the rate of 20 tons each, would fill 30,000 cars. At 20 cars to the train this would equal 1,500 trains, or a continuous train of cars 180 miles long. The agreement in area of this type of soil and the areas on the salt map showing 3 or more per cent of salt is almost identical. This type of soil being low, wet, salty, and of a clay nature, is not worthy of any notice for agricultural purposes at the present time. Those parts of it that extend far back into the better land form good drainage outlets for the latter and with very little improvement would serve to conduct drainage water to the lake. 100 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. .TORDAN MEADOWS. The Jordan Meadows comprise about 13 square miles, or 4.8 per cent of the entire area. They lie as a narrow strip, from a few rods to three-quarters of a mile in width, bordering on the Jordan River, and their usual elevation is only a few feet above the water in the river. Both the soil and the subsoil vary much in texture. The soil, how- ever, is generally either a sandy loam or a loam about 2 feet in depth, which is underlaid by 2 feet of clay and this in turn is underlaid by sand and gravel. The soil is usually black, on account of the large amount of organic matter that it contains. The following table shows the mechanical analyses of the soil from three places: Mechanical analyses of Jordan Meadows soils. 5= $ d d o i g a a ^^ a o o o H a No. Locality. Description. U si <0 •Sft 09 'o 1 .a O o Per w ia "in 1 ■oa 1 II 1 Per o d o d (» Per °a 2a 'So o d s g o d 3 Fer Per Per Per Per Per Per Per cent cent. cent cent. cent cent cent. cent cent. cent cent 433S W. C. sec. 22, T. 1 Eiver bottom. 1,0,'i 1.18 5.90 T. 3.71 8.4216.9136.07 15.75 1.74 9.65 N.,E.l W. 0to24inclies. 1 4337 SW. C. sec. 14, T. 1 S.,E.1-W. Eiver bottom, to 12 inches. 1.96 1.89 9.23 T. 1.57 1.72 7.16 31.30 26.76 3.34 15.51 4339 E. C. sec. 14, T. 3 S., E.1-W. River bottom, 0tol2inclies. 1.68 4.85 11.61 .21 .82 3.01 6. 87 19. 41 30.84 5.96 16.92 This type of soil was the first irrigated on the west side of the river, but wherever irrigation has been practiced above this laud the seepage waters have come down and caused much damage. At present very little of it is farmed, but it often furnishes good pasture. The accompanying table gives the salt content to a depth of 6 feet in three localities. The salt while not present in excessive amounts is sufficiently high to be harmful and is sometimes even fatal to ordinary crops. There is a tendency for it to accumulate at the surface. Salt content at saturation for varioiis depths and places in Jordan meadows. [Percentage calculated on water-free soil.] No. of boring. Depth in feet. Depth to standing ■water. 1. 2. 3. 4. 6. 6. 5 Per cent. 0.37 0.55 0.73 Per cent. 0.76 0.51 0.67 Per cent. 0.40 0.49 0.36 Per cent. 0,31 0.55 0.29 Per cent 0.36 0.35 0.27 Per cent. 0.35 0.30 0.18 Feet 6* 637 281 JORDAN SAND. These sands constitute a marginal area along the bluffs just above the Jordan meadows, and are found in a few isolated areas on the upland. They are shown on the soil map in yellow. It is fine sand Report No. 64, U. S. Dept. Agr. PLATE XIII. BINGHAM STONY LOAM — SALT LAKE SAND. 101 usually to a depth of 6 feet or more, although occasionally occurring as an overlying stratum only 1 or 2 feet in depth, covering sandy loam or loam. In places it drifts about as dunes, and is usually so located as to be difficult to irrigate, therefore it is not much used for agricultural purposes. Where irrigated it requires large amouuts of water to cover any considerable distance, and as a consequence, accumulations of seepage waters on low adjacent areas are the rule. Owing to the leachy character and good underdrainage, it seldom contains much salt. BINGHAM STONY LOAM. This loam occurs above all canals, and constitutes a small area near the foothills of the Oquirrh Mountains, shown on the soil map by the brown color. It consists of a thin layer of sandy or gravelly loam, under- laid by bowlders, rock, and conglomerate, which frequently outcrops, from a few inches to the height of a man, at the surface. It is too stony for cultivation. SALT LAKE SAND. This sand is a product of Great Salt Lake, and consists of spher- ules about the size of 'So. 10 shot, which are made up almost wholly of carbonate of lime. The accompanying table shows the mechanical analysis of a sample of this sand taken from the dunes along tlie lake shore. It is known as oolitic sand, because it resembles thepetriiied eggs of fish. Whether the spherules as such were formed in the proc- ess of separating from the water or whether they were formed by the action of the wind and water, from the broken fragments of lime-car- bonate hardpan that occur in great quantity along the shore, was at first a matter of conjecture. Mechanical analysis of Salt Lake sand. Diameter. MilliTneiers. 2 tol 1 to 0. 5 0.5 to .25 .25 to .1 .1 to .0001 Conventional name. Fine gravel Coarseeand Medinm sand Fine sand Very fine sand— silt and clay 4355. Shores of G. S. L. 0.62 2.35 81.25 15.32 .51 Upon examining the different separations in the laboratory, however, it was found that the portion classed as fine gravel consisted of quite angular broken fragments of lime carbonate, the angles slightly rounded by erosion. The particles of the next grade — coarse sand — were much more rounded, although the larger ones were still quite angular, as could be seen by the naked eye. The third grade — medium sand — included over four-flfths of the sample, and in this all of the particles were well rounded, there being no angular .ones. Under the micro- scope the most of these particles proved to be almost perfect spheres, while a smaller proportion of them were oblong or egg shaped. ™ The 102 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. surfaces were quite smooth and highly polished. The fourth grade — tine sand — was similar to the third, but more of the particles were oblong and a few of them somewhat cylindrical, while the small percentage remaining in the class of very fine sand was chiefly angular particles of very fine sand and silt with just a trace of clay, the material being silice- ous, not carbonate of lime. All of the particles, except those of the very flue sand, when treated with dilute hydrochloric acid give rise to a vio- lent ebullition of carbon dioxid and soon disappear, leaving behind only a small flocculent precipitate, which, under the microscope, is shown to be clay particles with an occasional angular particle of silt. There seemed to be no nuclei to the lime carbonate particles, and all the evidence points to their formation by the breaking up of the thin pieces of lime-carbonate hardpan, which is quite abundant. These broken pieces, when sufficiently cubical, afterwards become rounded by the action ot the water and the wind. Where the particles of lime- carbon- ate hardpan are flat they do not take, on a rolling motion, and conse- quently do not become rounded. The particles are sufflciently soft to be easily crushed by pressure with a knife blade. It forms an insignifi- cant area along the shore, either as dunes, which in some places reach 10 or 12 feet in height, or spread out on the beach and on the slightly ele- vated areas near the lake shore as a layer of from a few inches to sev- eral feet in depth. Whatever its mode of formation, the material evi- dently comes from the lake water, which has reached the saturation point in regard to carbonate of lime. HARDPAN. The sketch map of the valley on page 103 shows a number of small areas of hardpan on the level lands between Salt Lake City and Great Salt Lake. This formation occurs over an aggregate area of about 12 square miles. The hardpan usually occurs under the Jordan sandy loam. This material is encountered at from 12 to 30 inches below the surface and at an average depth of 18 inches. It is from 2 to 18 inches thick, and averages about 3 or 4 inches. The texture of the hardpan is the same as the material immediately above and below it, but this layer has been cemented by lime carbo- nate. Under ordinary conditions it is quite pervious to water and to the roots of plants, but when dry it is quite hard and dififtcult to dig. When moistened with water and soaked for a while it softens consider- ably, but does not disintegrate to any appreciable extent. It effer- vesces freely with hydrochloric acid and falls apart into a sandy loam. As would be expected, the subsoil immediately below the hardpan is quite moist throughout the season, while above it the soil is quite dry during the summer months. The soil above the hardpan is usu- ally free from excessive quantities of alkali, while below the hardpan the salt content is very much greater. On the shores of the Great Salt Lake very interesting observations Report No. 64, U. S. Dept. Agr. Plate XIV. 05 H O r^ ° 2 > f^ > m m o o "" "n H > r a fi ^ 4'* 7i r'.^ife- '>/ I V ^..W / ti SKETCH MAP OF SALT LAKE VALLEY, UTAH. 103 Mote Fi gures show parts of soluble matter in 100,000 parts of water Shaded areas shour hard-pan 1~5 feet belozc surface A-A \ B-B \ Sect Ions c-ci milCs ^ . 3 JA/. I fi . 2 \A/. /? . / i^ . Fig. 10. — Sketch map of western part of Salt Lake Valley, showing canals, hardpan areas, and parts of salt in 100,000 parts of river and irrigation waters. 104 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. were made on hardpan forming on the surface and in all stages of for- mation. The preliminary stage was the washing up of algse from the lake and of immense quantities of the salt shrimp, which lives in the lake waters, on to the beach. These gradually become crusted over with various salts,' which, after repeated leaching, leave only the difft- cultly soluble lime carbonate. As the organic matter decomposes it is replaced with the lime carbonate, until the mass finally assumes the character of thin strata of broken limestone rock, nearly covering the surface over considerable areas. Dr. Cameron has examined this material and has made the follow- ing interesting statements upon its probable mode of formation: The samples collected on the shores of the lake hy Mr. Gardner suggest some very interesting ideas as to the formation of hardpan. The samples all show the presence of calcium carbonate as such, more or less densely compacted, and there can be no doubt that this material is the cement holding together the other constituents which go to make up the hardpan. Calcium, in some form of combination, is present in the water of the Salt Lake. From an analysis made by Dr. J. E. Talmage in 1889 there appeared to be 0.084 per cent of calcium in the water, and it seems reasonable to sup- pose the amount is no less at the present time. The original source of the lime was probably in gypsum or carbonate of lime, and for our purpose it is a matter of indifference which it was. Gypsum (or the carbonate of lime either) is but slightly soluble in water free from other salts. But the presence of another salt markedly influences its solubility. If sodium chlorid be present, we shall have in the solution some sodium chlorid, some sodium ions, and some chlorin ions, thus : + - NaCl Na 01 When the calcium sulphate is brought into contact with the solution we will get at once and in the same way calcium sulphate, calcium ions, and sulphions in the solution : + - CaSO, Ca S0^ But the sodium ions and the sulphions present will unite to a certain extent to form sodium sulphate. At once there is a tendency for the sodium chlorid to furnish more sodium ions and the calcium sulphate to furnish more sulphions to restore the equilib- rium. In the same manner the calcium ions and the chlorin ions will unite to form the very soluble calcium chlorid in the solution with a corresponding tendency to the formation of further calcium and chlorin ions. While the gypsum is but spar- ingly soluble, all these other salts formed are quite soluble, and more and more of the gypsum will dissociate and go into solution until equilibrium be reached. If the sodium-ohlorid solution be at all concentrated, it has been shown that the amount of gypsum soluble in it is astonishingly large. In this way it is easy to see how lime salts could be carried down to the lake. But just as the presence of a salt increases the solubility of another without a common ion, it decreases the solubility of the second salt if they both yield a common ion. Of all the possible salts which may form in a solution when in instable equilibrium as regard saturation, that salt which is the least soluble will, of course, be precipitated. Thus it is that the waters of the Salt Lake containing large amounts of soluble sulphates, practically saturated with respect to them in fact, contain but a very small amount of calcium. Moreover, the water of Salt Lake has been shown in this laboratory to contain a small amount of carbonate, either calcium carbonate or, much more likely, sodium carbonate. This fact is not recorded in any of the published analyses of the water, so far as I know, and as a matter of fact this need not cause any surprise under the Report No. 64, U. S. Dept. Agr. Plate XV. HARDPAN. 105 circumstances. Its presence was discovered almost accidentally. Mr. Gardner had attempted to test the water for alkalinity by adding phenolphthaleiu solution. No color whatever was apparent, but on throwing the contents of the vessel away and attempting to rinsB with distilled water a marked alkaline reaction was observed in the wash waters. Investigation showed the absence of any alkali in the distilled water. An examination in this laboratory brought out the explanation of the phenomena very clearly. There is sodium carbonate in the solution, and normally this would dissociate with the formation of sodium ions, which could be detected by the phenolphthaleiu. But the water contains so large an amount of salts with sodium ions that the solution is saturated with respect to this ion, and, carbonates being salts of a weak acid — that is, with a relatively small tendency to dissociate — the dissociation of the sodium carbonate is completely " driven back.'' Consequently there is no dissociated sodium carbonate in the solution. On the addition of more water, however, the concentration of the solution with respect to the sodium ion is decreased, more sodium ions may be formed, and the sodium carbonate dissociates, which fact is indicated by the phenolphthaleiu solution. This explanation was verified by repeated experiments with the water from the Salt Lake and with brines containing traces of sodium carbonate, prepared artificially. There is the further interest in these facts, that the absence of lime carbonates in the lake water has been the subject of much interest and speculation. The explanation is apparent from what has just been said. As the conditions are such that practically no disso- ciated calcium sulphate or dissociated calcium carbonate can exist, the water will only take up so much of these salts as is soluble without dissociation, somewhat less, in fact, than if the other salts were not present in the water. At the edge of the lake the spray carrying various salts in solution falls on bunches of algse and other organic matter, and, the water evaporating, the salts contained are precipitated. Theoretically the first to separate should be the small amount of calcium carbonate present, and in a general way this is probably so, but it is more or less mixed with the other salts. These other salts, being more soluble, are partially restored to the lake by the returning drip, by washing with rain water, etc., and the calcium carbonate already deposited will hasten the precipitation of more calcium carbonate from successive washings of spray, and thus it will gradually accumulate. In the specimens of the most recent formation collected by Mr. Gardner these views are well exemplified. In water, either cold or hot, the material is partially disintegrated and dissolved, fragments of the algse (green and well preserved) being obtainable from the residue, as well as small twigs and other organic matter. The contents of the water solution appears to be in the main sodium ohlorid, though other salts might be found on a careful examination. The solution showed the presence of but little lime and did not effervesce noticeably on the addition of an acid. The residue insoluble in water proved to be nearly entirely carbonate of lime, a very small residue being left after treatment with dilute hydrochloric acid. Farther back from the water's edge specimens were obtained, consisting of twigs, grasses, etc., dead organic matter, bound together with a material which j)roved on examination to be nearly all calcium carbonate, though there appeared to be quite a small amount of sodium chlorid and other salts present. On treatment with dilute hydrochloric acid, besides the dead organic matter there was a small residue of fine sand or silt left. This hardpan, while much harder and more compact than the one first described, is still sufficiently friable to be broken with the fingers, though with some difficulty, and is fairly porous. It can be regarded as representing a subse- quent stage or later development of the first described specimens. The sodium chlorid and other water-soluble salts have been gradually washed out and returned to the lake. The calcium carbonate, by alternate redissolving and reprecipitation, has become more dense and compact. In this process the sodium ohlorid may have been an important factor. It would seem quite probable that the dead organic mat- ter might play a decided r61e in yielding by oxidation carbonic acid. This carbonic 106 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. acid, in tlie presence of moisture, would dissolve the lime carbonate with the forma- tion of the bicarbonate, and the bicarbonate, in turn, would reprecipitate as the car- bonate on drying. Still farther back from the water's edge was found another type of material, evidently representing a later stage in the genesis of hardpan. This material appears to be but very slightly affected by treatment with water. It is practically all soluble in moderately dilute hydrochloric acid, a small residue of light brown silt remaining. The cementing material proves to be almost entirely lime carbonate with, perhaps, some magnesium carbonate. It is quite dense and compact and noticeably free from the undecomposed organic matter in the speci- mens last described. It evidently represents a later product of the process of resolu- tion in carbonated water, reprecipitation on drying, and may fairly be taken as the type of the final product in the genesis of a lime carbonate hardpan. On the shore between the two deposits last described (or still farther back in the form of wind-blown dunes) are found great quantities of small, remarkably well- rounded spherules, whose composition appears to be either entirely calcium car- bonate or calcium carbonate about a very small nucleus of sand or some siliceous mineral. Several hypotheses have been suggested as to the origin of this material, but so far none seems to have received much credence. I am inclined to think they originated from the forming hardpan on the shore, by the action of waves during rough weather, by splitting up of the parent material by frost, the result of wind action, or of all these agencies. Their rounded form is quite easily accounted for in two ways — as the result of rolling from the action of the water; by the etching due to the solvent effect of the water, it being a well demonstrated fact that in the absence of any special reason to the contrary the result of such action is always to round oft' the edges and produce a spherulitic form, thereby reducing the surface, and consequently the "active mass," to a minimum. The striking uniformity in the size of the particles may be taken as ooniirmatory evidence of the views just pre- sented. Suppose the particles to have been originally of widely varying masses. At such times as they would have been under the influence of the water they would have been stirring about in a heavy, rather viscous brine, whose specific gravity would not be very far from that of the carbonate particles. Naturally the hea /ier particles would gradually accumulate at the bottom, the lighter ones at the top, iind, as the turbulent actions ceased because the particles had practically the same shape and density, those of equal mass would have equal volume and would settle at the same time. The larger particles at the bottom might well be expected to gradually consolidate if left undisturbed for suiBcient time, and some specimens collected by Mr. Gardner indicate that such action has taken place. WATEK SUPPLY. The main water supply for the western part of the Salt Lake Valley is derived directly from the Utah Lake and delivered through the Jor- dan Eiver, the distance between the lake and the Jordan Narrows, where the later canals are taken out, being about 6 or 7 miles. Utah Lake is a large body of fresh water, having approximately 125 square miles of surface, but rather shallow. It is fed by rather short mountain streams, derived largely from melting snow. The water in these streams is of excellent quality and quite free from salts. The water of the lake itself contains rather more alkali, as the seepage from the surrounding lands — both irrigated and nonirrigated — sensibly affects the salt content. In the spring of 1899 Mr. Means found the salt con- tent of Utah Lake to be about 50 parts in 100,000. In a shallow lake with such an extensive surface area the effect of evaporation during the summer must sensibly increase the concentration of the salts in the WATER SUPPLY. 107 water. This probably accounts for the difference of the salt content of the lake, as observed by Mr. Means in July, and the 89 parts per 100,000 at the N"arrows in the Jordan Eiver, observed aboat the 1st of October by the writers. There are seven canals taken out of the Jordan River between the Jordan Narrows and Salt Lake City for irrigation purposes, besides three or four small canals for water power. Five of the irrigating canals are on the west side of the river, but only three of these are at present used to any considerable extent for irrigating purposes. The total flow of water in the Jordan Eiver at the Narrows has been variously estimated at different periods between 1895 and 1899 at from 244.2 second-feet to 526.3 second-feet, the information being obtained from the of&ce of the city engineer. The North Jordan Canal was the first one of the great canals con- structed on the west side of the river. It is taken out of the river about 9 miles below the Jordan Narrows. McAllister made a number of measurements in 1895-1897, inclusive, and found the flow in the canal to vary from 50.3 to 103.7 second-feet. This canal is about 14 miles long. The next canal constructed on the west side of the river was the South Jordan, which is taken out of the Narrows, and is about 16 miles long. Various measurements have been made by a number of observers from 1895 to 1899 of the flow in this canal, and the results vary from 40.4 to 166.6 second-feet, with an average flow of about 75 or 80 second- feet. The Utah and Salt Lake Canal, which is about 50 feet above the last-named canal, is also taken out of the river at the Narrows and is about 23 miles long. The measurements in this canal at dift'erent times and by different observers vary from 49.8 to 185 second feet, with an average of considerably more than 100 second-feet. A con- siderable part of this, however, amounting at times to 50 per cent, is used by the power plant which supplies an electric current to Bingham and Mercer and returns the water directly to the river. Observations made in 1899 show an estimated flow on June 4 for all canals taken out of the Jordan Eiver of 526.3 second-feet; on August 18 of 325.6 second-feet; and on September 22 of 325.8 second-feet. Besides the three principal canals used for irrigation on the west side, there are two others, namely, the Brighton and North Point and the North Point Consolidated, which were intended for irrigating the low land west of Salt Lake City, but which are hardly used at all at the present time. The North Point Consolidated Canal has a capacity of about 100 second-feet; the Brighton and North Point Canal is much smaller. There was enough water flowing past the intake of these canals during the summer of 1899 to have supplied their full capacity. An analysis was made by Dr. Cameron of the solid matter of a sam- ple of water from the Jordan Eiver near the Narrows and of another 108 A SOIL SURVET IN SALT LAKE VALLEY, UTAH. sample taken from the Jordan River opposite Salt Lake City. The fol- lowing table gives the result of these analyses: Kiyid and amount of salt in Jordan Birer water aa determined by Dr. Cameron. At intake of Utali and Salt Lake Canal, Oc- tober 2. At Salt Lake City, October 30. Kind of salt. Parts of salt in 100,000 parts of water. Per cent of total salts present. Parts of salt in 100,000 parts of water. Per cent of total salts present. NajCOa 4.2 23 11 5.4 45.6 4.71 25.78 12.33 6.06 51.12 Trace. 38 9.8 6.4 55.8 CaS04 34.86 Na.SOj ■. 8.99 M"CI, 4 96 NaCl 51 19 Total 89.2 98 8.8 100 109 118.4 9.4 100 Eesidue dried at 105° C Water of cryatallization by difference The sketch map of the valley (page 103) gives the total salt content in parts per 100,000 as observed during the progress of the survey in difi'erent partsof the valley. It will be observed that there is considerablefluctua- tion iu the salt content of the river between the Narrows and Salt Lake Oity. This is dependent on two causes. On the east side of the river there are a number of mountain streams which deliver a considerable volume of snow water to the Jordan River, thus improving the quality of the water by dilution. On the other hand, there is a large amouut of seepage waters collected by the river from the lands on either side, and, as these waters are generally highly charged with alkali salts, the salt content of the river tends to increase. Without this explanation of the conditions prevailing it would seem strange to see a salt content at the Narrows of 89 parts per 100,000 and this increased, at a point about half way to Salt Lake Oity, to 197 parts; at the intake of the surplus canal it is 149 parts, while at a point only two or three miles from the mouth of the river the salt content is 126 parts per 100,000. The increase below the Narrows is to be ascribed entirely to the seepage from the adjoining irrigated lands on either side, while the comparatively low salt content near the mouth is due to the large vol- ume of purer mountain water delivered by the several creeks which empty into the Jordan River at Salt Lake City. The sketch map shows the salt content of the water in the irrigation canals so far as observed during the course of the survey. These meas- urements cover a period of four mouths. For plants growing with their roots immersed in water, i. e., grown by water culture and not in the soil, the limit of endurance is about 1 per cent, or 1,000 parts of salt per 100,000 parts of water. A soil con- taining 0.4 per cent of salt, saturated with pure distilled water, would have in the soil moisture a concentration of 1 jjer cent. As all of these soils contain more or less alkali, and as evaporation and consequent Report No. 64, U. S. Dept. Agr. Plate XVI. APPLICATION OF WATER. 109 coucentration set in immediately after the application of water to the land, it is unsafe to use water having a coucentration greater than about 250 or 300 parts per 100,000 for irrigating the lands. According to this standard it will be seen that all of the canals de- liver water of a good quality for irrigation purposes, especially the higher canals. The Utah and Salt Lake Canal contains only about 95 parts of salt per 100,000 parts of water, the South Jordan Canal con- tains hardly more than this, while the North Jordan Canal contaius nearly twice as much, but is still well within the limit of safety, at least so far as immediate effects are concerned. The limit of safety is depend- ent upon so many things — such as the salt content of the soil, the tex- ture of the soil, the drainage, kind of crop, the stage and condition of cultivation, and the climatic conditions — that only very general figures can be given for such broad application. APPLICATION OP WATER. The three principal canals on the west side of the river with which we have to deal are the joint property of the owners of the irrigated land, each man having shares in proportion to the amount of land owned. Anyone not holding shares can rent water rights from those who own more shares than they have personal need of. The water is generally apportioned among the landowners in proportion to the stock they control. The exact amount of water used per acre in this district has not been determined; but the average for the State of Utah is esti- mated at about 1 second-foot for each 100 acres. There is generally an abundance of water in the canal; but when there is any deficiency all suffer alike in a reduced supply. The water is allowed to run con- tinually throughout the season and the excess runs onto uncultivated or pasture land. There is believed to be more water used at times of abundant flow than is absolutely necessary. Furthermore, the high- land canals run for a portion of the distance through very pervious gravelly loams. The seepage and waste waters from the canals account in great measure for the 10 square miles of good land which has already been ruined by seepage and alkali. It has been shown that the water is of good quality and the lands of the upper benches are naturally free from any great excess of alkali; but the continual seepage from the canals during the growing season for a great many years has transported a quantity of salt to the lower levels. The necessity of careful construction of the canals, especially those on gravelly lands, and the desirability of preventing the waste water from flowing over the lower levels is sufQciently obvious without further comment. The application of water on the low lands west of Salt Lake City, where there is a large amount of alkali in the lower depth, has been attended with very disastrous results to crops. The salt has quickly 110 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. risen to the surface and, even where the surface foot was originally free from alkali, the crops have been completely ruined in the course of two or three years. A very serious feature of the prevailing practice is that the land upon which an excess of water is used, or the land adjacent to a leaking ditch, is often not injured for a while and may even be improved by the excess of seepage water; while the lands at the lower levels, perhaps under the second or third ditch, may receive the full effect of this pernicious practice or condition. The advisability and even necessity of State legislation, to compel the ditch owners to guard against undue seepage and 1o prevent the property owners from using excessive amounts of water in irrigation, is sufficiently obvious to require no further comment at this place. Property owners whose lands are damaged by either of these means should be able to recover damages in civil suits. UNDERGROUND W^ATBR. One of the maps accompanying this report gives the depth to stand- ing water at the time of the survey. One of the three shades of green shows the area upon which standing water is found within 3 feet of the surface of the ground; another shows the area where the water is between 3 and 10 feet of the surface, and the third color shows where it is below 10 feet. With standing water within 3 feet of the surface of the soil it is in no condition for any of our agricultural crops. Such areas must be drained in order to use the land for agricultural purposes. With water within 3 to 10 feet of the surface there is always more or less danger in the application of irrigation water. It is incumbent upon the owef s of such land to watch the fluctuation of the underground water and to under- drain wherever necessary. The texture of the soil, and especially the character of the subsoil, will largely determine the extent of the danger to be feared. The large areas south of Salt Lake City and the areas along the Jordan Eiver with standing water within 3 feet of the surface are believed to be due entirely to seepage from the canals and from the irrigated lands above. The underground water map shows, therefore, the areas which require immediate drainage to relieve the soil of surplus water. It is incum- bent upon all those living within the areas colored medium and dark green to provide adequate drainage when the danger from underground water becomes imminent. One of the pernicious effects of the accumulation of seepage waters is that the soil is made closer and more difficult to drain, therefore arti- ficial drainage can be more economically and effectively applied before, rather than after, the collection of seepage waters near the surface. Report No. 64, U. S Dept. Agr. Plate XVII. r > CO H in I B > %*' A V^; ' 'I I 4* s .^ */"• . PI ** '^ V . - i - ^ i . i •I r •" ' i f " *.' ALKALI IN SOILS. Ill ALKALI IN SOILS. The following table gives the composition of the alkali in the soils and crusts from a number of localities, as determined by Dr. Cameron. It will be seen that the sodium chlorid constitutes from 50 to 97 per cent of the total salts. The next largest constituent is sodium sulphate. The calcium sulphate is a difficultly soluble salt, and when the water evaporates this will be deposited in the soil as harmless gypsum which will not readily go into solution again. The other salts are all quite soluble and are liable to accumulate and concentrate in the soil mois- ture upon evaporation of the water from the surface of the land. Chemical composition of salts in crusts and soils as analyzed by Dr. Cameron. No. Locality. FajCOa, CaCla MgCU. CaSO,. MgSOi Na5S04 NaCl. 4366 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 S. 1, T. IS., K. ] W 3 miles northwest of S. L. S.32, T.1N.,E.1W S. 14, T.3S., E. IW S. 24, T. 1 S., E. 2 AV S. 2, T. IS., E. 1 W S. 16, T. IS., E. 3W S. 29, T. IS., E. IW Chambers, 2 miles east — S. 4, T. IS., E. IW S. 1, T. IS., E.2 W S. 14, T. 3S., E. 1 W 0.96 9.28 0.77 0.38 Trace. Trace. Trace. .23 10.41 .30 Trace. 6.21 'io.'sf "4.77 0.25 .37 3.16 .12 29.87 .75 22.61 31.15 43.12 11.97 8.29 .71 2.64 32.32 9.75 49.' 43 96.90 66.84 46.17 73.18 85.94 55.48 89.31 59.98 ; 72. 04 49.30 83.26 54.49 The amount of alkali was determined by the electrical method in every foot down to at least 6 feet in depth. The alkali map represents what may be considered the average conditions, according to the judg- ment of the observer. If a boring showed a small amount of alkali in the surface foot and a large amount in the remaining depth the soil was classed as unfit for cultivation, as it would require but one or two applications of water to bring an excessive amount of salt to the surface. Attention is called to the very large accumulation (over 3 per cent) of soluble salt in certain areas in the land west of Salt Lake City. There is a general agreement between the alkali map and the soil map, as would be expected, but this is influenced to a considerable extent by the topography of the country. The heavy clay soils. and the land having this material within a short distance of the surface have gen. erally the largest accumulation of alkali, on account of the imperfect drainage. The soil of the uplands is naturally free from excessive amounts of alkali, while the level area between Salt Lake City and the Great Salt Lake contains excessive amounts of salt. This latter fact is undoubtedly due to the influence of the Great Salt Lake, which, within comparatively recent years, covered much of this area. There are no antidotes for this kind of alkali, with the exception of the sodium car- bonate, and adequate artificial drainage is the only practical means of reclaiming the land and providing against further disaster. The pos- sibilities of reclaiming the level tract west of Salt Lake City will be deeerib^d under ei subsequent head. 112 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. Oue interesting fact brought out in this investigation is that alfalfa appears to stand a slightly higher salt content in the Salt Lake Valley than either in the Yellowstone Valley of Montana or the Pecos Valley of New Mexico. This may be due to the longer period in which agri- culture has been practiced in this locality and the gradual adaptation of the alfalfa to these alkali lands. This is a matter which requires fuller investigation by the Vegetable Physiologist. It is clearly apparent, from the investigations in the Salt Lake Valley, that our different staple crops can stand different amounts of alkali. These limits have been shown on the alkali map by different colors. A number of factors enter into the question of the limit of endurance of plants for alkali. With the same amount of alkali plants will suffer less in the heavier soils than in the sandy. They will stand more alkali with thorough cultivation, and they will often stand a considerable amount of alkali if they are started under favorable conditions. In some districts of California sugar beets do well on soils containing a large amount of alkali. They are planted in the spring when the ground is wet either by rains or by previous irrigations, which carry the alkali into depths of the soil. When the soil dries out the alkali is brought to the surface and is left above the area of the active roots. It is a common practice in some localities to irrigate heavily just before planting in order to accomplish this very purpose. Obviously this method would not be successful when the soil to a considerable depth contains an excessive amount of salt. Along the ridges and draws in the level west of Salt Lake City, where good drainage is secured, crops are frequently cultivated with a moderate degree of success when the land has a salt content higher than would be permissible were the drainage less complete. In a con- siderable percentage of this area the surface is comparatively free from alkali. Many attempts have been made to bring such land under irrigation, but the results have been disastrous after one or two years. In the upland soils of this locality the excessive accumulation of alkali, in land which was formerly free from salt, is almost invariably preceded by an accumulation of seepage water. The treatment for alkali, therefore, in soils previously free from salt, is almost always accompanied by the problem of getting rid of the seepage waters, BLACK ALKALI. The corrosive sodium carbonate is present in considerable amounts in the soils west of Salt Lake City. The best antidote for this, as Hilgard has pointed out, is the application of calcium sulphate, which under proper conditions of drainage and aeration, converts the sodium carbonate into sodium sulphate or white alkali. On every area in which there is an excess of sodium carbonate there is also an excess of white alkali. It would be a waste of money, therefore, to apply calcium Report No 64, U, S. Dept. Agr, Plate XX. CO > £' o s > UNDERDRAINAGE AND THE RECLAMATION OF WASTE LAND. 113 sulphate, as the sodium carbonate would be washed out along with the white alkali on the introduction of proper drainage and flooding. The limit of endurance of plants for sodium carbonate is assumed to be that which Hilgard determines for the California soils, namely, 0.1 per cent. The lower limit of what may be called the danger line has been placed at one-half this amount. Attention is called to the very large amount (over 0.25 per cent) of sodium carbonate in the soils of certain areas. The surface crusts occasionally contain as much as 10 per cent of sodium carbonate. Quite often there are large accumulations of calcium chlorid, which in one instance has amounted to about 40 per cent of the total salts present. There is also frequently an appreciable amount of strontium chlorid in the crust, and the suggestion is made that some of the deposits may be sufflciently rich in one or more of these three salts to warrant commercial development. No special attention, however, was paid to this particular feature. UNDERDRAINAGE AND THE RECLAMATION OP AVASTB LAND. Attention has already been called to the necessity of underdrainage for protection against injury from seepage waters and alkali and for the reclamation of injured lands. Irrigated lands in the Salt Lake Valley are worth at least from $60 to $100 per acre. Lands immedi- ately adjacent to Salt Lake City, especially if held as suburban prop- erty and if free from alkali, would be worth much more than this. There is plenty of good tile clay in the vicinity of Salt Lake City, and tile could be manufactured for the farmer at a reasonable cost. It is estimated that it would cost from $10 to $20 an acre to underdrain these lands which, under the present conditions, have a merely nomi- nal value. Lands in New York, Ohio, and Illinois, worth from $50 to $75 per acre, have been very extensively underdrained in order to increase their productiveness, to hasten the maturity of the crops, and to insure the crops from inj ury by drought. I fc would certainly be a reasonable prop- osition to protect these valuable lands and to reclaim in the same way what would be valuable land. Money so invested is in the nature of an insurance against loss of crops from seepage waters and alkalL During the course of this investigation particular attention was given to the possibility of reclaiming the vast tract of 125 square miles between Salt Lake City and the Great Salt Lake. The levels of the railroad surveys and of the canal companies were freely consulted. At Salt Lake City the level of the Jordan Eiver is about 20 feet above the level of the water in the Great Salt Lake. The distance across is about 14 miles. There is a slight ridge, however, running a little west of north, about a third of the way across from Salt Lake City. Prom the crest of this ridge to the Great Salt Lake there is a uniform fall of approximately 3 feet to the mile. This would be ample for the main H. Doc. 399 8 114 A SOIL SURVEY IN SALT LAKE VALLEY, UTAH. drainage canals, as the irrigating canals have only about one half this fall. Furthermore, there are many draws, already 4 to 8 feet deep, extending like lingers through this area, which with little additional work could be made to answer for a considerable part of the drainage system. On account of the impervious nature of the Jordan clay, the great salt content, and the low elevation, it would not be advisable to attempt drainage over this class of laud at the present time. Subtracting this area, estimated at 35 square miles, from the 125 square miles, the value of the remaining lands, if thoroughly drained, would be about $3,000,- 000. At present they have merely a nominal value. Any large drainage system of this kind can be established more efficiently and economically by a company than by individual effort. For an enterprise of such magnitude — so nearly affecting the welfare of a large number of people — the State or county could well use its credit in assisting the undertaking. This is commonly done in similar enterprises in other localities. There seems to be little doubt of the feasibility of reclaiming this land from the engineering point of view and, with the abundant supply of water, there is still less doubt of the efficiency of the system when once introduced. The possibilities should appeal to the commercial spirit of the jieople and induce capital to undertake this very desirable enterprise. ACKNOWLEDGMENT. The authors wish to acknowledge their indebtedness to Col. 0. A. Stevenson, secretary of the Utah State Irrigation Society, and to Messrs. George A. Lowe, T. J. Almy, E. E. Jerome, and others for their assist- ance and interest in the investigations. Also to Mr. Wilks, county surveyor, and Mr. Kelsey, city engineer, for base maps and data relating ti> water measurements and levelsforvariousplacesin Salt Lake County. Officials of the Oregon Short Line liailroad kindly furnished transpor- tation over all of their lines within the State, which was of material assistance. A RECONNOISSANCE IN SANPETE, CACHE, AND UTAH COUNTIES, UTAH. By THOS. H. MEANS. GEOGEAPHY AND TOPOGRAPHY. A portion of July, 1899, was spent in a rapid survey of the alkali soils of Utah, Sanpete, and Cache counties in Utah, with the following results, briefly given. The conditions in these three districts are in general very similar, yet each district differs from the others in certain respects. Utah County lies directly around the fresh-water lake that bears the same name, and all of its irrigated land is on what was, at a previous epoch, the bed of Lake Bonneville. The soils are the sediments of this ancient lake, more recently- modified by the inilowing streams and by weathering. The irrigation water supply is obtained from the mountain streams, the principal ones being American Fork, Provo Eiver, and Spanish Fork. The source of the water supply for these streams, during the irrigation season, is chiefly the melting snow on the high mountains of the Wasatch Range, and the water is therefore of excellent quality. It contains about 15 parts of soluble matter per 100,000. The irrigation systems consist of a series of canals and ditches owned and operated by the farmers. Parts of the valley have been under cul- tivation for about forty years. When the valley was first settled the strip of land immediately around Utah Lake was wet and swampy. By irrigating the upland the wet land has greatly increased in area, this increased wet area becoming less valuable, although still furnishing desirable grazing. The principal damage is from seepage waters, but where the soil is heavy, salts have accumulated at the surface in sufiQ- cient quantities to seriously damage the meadows. Moderate amounts of salts are everywhere present in the soil and through the agency of irrigation and seepage waters they are concen- trated within small areas. An examination of the seepage waters from the benches showed comparatively small amounts of salts present, but these mildly charged solutions are often much concentrated by evapora- tion and give rise to large accumulations of salts in the soil. 115 116 RECONNOISSANCE IN UTAH. Sanpete County is located in the central part of Utah and its farming lands lie in the valleys of the San Pitch and Sevier Rivers. The land lies at a greater altitude than the ancient Lake Bonneville, and the soil formation is therefore somewhat different from that of the preceding county. The valleys are more level and slope more gradually to the mountains than in the lake basin proper. The irrigation water supply is derived from several sources. Around and above Manti the water is taken directly from the mountain canyons and is good. Ten miles above Gunnison the San Pitch is dammed, thus forming a storage reservoir which supplies water farther down the valley, around the junction of the San Pitch and the Sevier. The water of this reservoir, at the time it was examined, carried between 50 and 100 parts of soluble matter per 100,000. The lands to the south of Gunnison receive their water supply from the Sevier River, as do also the lands around Richfield. The water of the Sevier River during a part of the year is diverted at a point above Gunnison. Grad- ually a portion of it returns to the river as seepage and at Fayette it is again all taken out for irrigation. Still farther down, at Deseret, the stream is drained for the third time. The water as first taken from the river is good; but the second time it is contaminated by the seepage from the Gunnison district. The water in the Fayette Canal contained, in July, 1894, 250 parts of soluble matter per 100,000; while in the Sevier River, just below Fayette, it contained over 300 parts. At Deseret it was probably still worse. The application of water containing more than 300 parts of soluble matter per 100,000 is attended with great risk unless precautions are taken to prevent concentration within the soil. Cache County lies in the north-central partof Utah, the farming lauds receiving water directly from the tributaries of the Bear River. The irrigation water supply is taken from the mountain streams and is of excellent quality. SOILS. The soils of the three districts mentioned above are so similar that they may be conveniently discussed under one head. The farming lands of both Cache and Utah counties are all in the Bonneville beds, and the soils were formed from sediments deposited from the ancient lake, and have, since its subsidence, been considerably modified by inflowing streams and by weathering. The soils of Sanpete County lie above the level of the ancient lake and were formed by material brought down from the adjacent mountains. They are very similar, however, to those of the two preceding counties. In each of the districts the soils in several parts vary from light sand and small gravel, through all grades, to the heaviest and most tena- cious clays. The upper benches of the deltas around the mouths of the canyons, and also the shore benches of Lake Bonneville, are covered with gravelly soils, grading down into coarse gravel. These soils are ALKALI m SOILS. 117 well underdrained and therefore free from salts, but owing to the thin- ness of the soil proper, and to the difliculty of applying water and cul- tivating the soil, they are but little farmed. The soils of the lower benches contain less gravel, but are sandy and of light texture. Upon the lower and more level parts of these valleys there are great variations in the soils. In those parts farthest fron< the iuiiowing streams, where the water movement is slow, the soils are heavy and often contain as high as 50 per cent of clay. As we approach the entrance of streams, where the water movement becomes more and more rapid, the soils are noticeably lighter, grading through loam, sandy loam, and at the mouths of streams becoming sandy or gravelly. Irrigation on the loose soils results in the transportation of considera- ble salt to the lower and heavier soils, where it is most difficult to get rid of. Around Utah Lake there is considerable of this heavy land that has been more or less damaged by seepage and alkali from the lands above, and in Cache Valley tliere is a large area of wet clay land which is used for meadow and grazing. Though of considerable value in this way, the value could be much enhanced by drainage and cultivation. ALKALI IN- SOILS. In Utah and Cache counties both the black and white forms of alkali are present. Origiaally moderate amounts of alkali were prob- ably everywhere present in the soil, and through the agency of irriga- tion and seepage waters a part of these salts has been transported and concentrated in local spots. The three following analyses, by Dr. Cameron, show the approximate composition of alkali from Utah County : 4161. 4162. 4160. Black White Crust from crust, 2 crust, 3 shore of miles miles Utah Lake. H.W. of If.W. of Provo. Provo. CaSOi 3.96 33.68 MgSOi Na-jSOi 10.02 NaCI 52.34 79.09 26.69 IfajCOs 0.00 20.91 1.26 The crust, No. 4160, was collected from near the shore of Utah Lake upon a bar surrounded by swamps. The water of the lake contains gypsum in solution. Gypsum acts as a chemical correction for the black alkali, changing it over into sodium sulphate. Here the process has been complete; all of the sodium carbonate has disappeared. In sample No. 4162 the per cent of sodium carbonate is so small as not to show black in the crust as collected. The black alkali, where present, is in spots and small areas and its reclamation with gypsum will prove an easy matter. 118 RECONNOISSANCE IN UTAH. In Cache County the black alkali is limited to spots and tracts in the west-central part of the valley and is always associated with the white alkali. Gypsum is found in some of the heavier soils and serves a valuable office in ameliorating the severity of the black alkali. The following analyses, by Dr. Cameron, show the composition of the salts under the two typical conditions : 4172. Tule swamp crust, 5 miles SW, of Lopan. «74. Black crust, 3 miles SW. of Logau. CaSO. .. MgSOa . Na,S04 . NaCl... Na^COa . 46.38 41.06 12.56 In Sanpete County the alkali, so far as examined, was all white. The presence of gypsum in nearly all of the soils and in the irrigation waters would lead one to expect this. The following analyses, by Dr. Cameron, illustrate the variation in the composition of the salts : 4165. Bed crust, 5 milea S. of Gunni- son. 4166. White crust. 5 miles S. of Gunni- son. 4170.- Yellow crust from kaolin mine, 2 miles SE. of Gunni- son. 4171. White crust un- der loose soil from hillside, 3 miles SE. of Gunni- son. CaSOj . NaSOi . MgSO^. MgClj . NaCl -. 1.27 73.40 3.14 89.03 25.33 7.83 7.18 4.47 85.33 3.51 95.60 Samples 4165 and 4166 represent the alkali as found in the soil, while 4170 and 4171 were collected to represent some of the possible sources of the salts. In each of these samples gypsum is present in sufficient quantity to prevent the formation of black alkali in a well-aerated soil. In the foothills south of Gunnison rock salt is mined; the streams cutting through these hills are likely to become contaminated with this salt and carry it to the irrigated lands below, much to their detriment. No actual exposure of the salt is at present visible along the stream beds, but the underground flows and heavy floods, no doubt, carry much of this salt valleyward. One of the highland canals, running along the base of these hills, leaked in a gravelly place and the resulting seepa.ge at the foot of the hill was heavily charged with salt and formed a heavy crust over the surface of the ground. There is much apprehension that this upland canal, through its seepage, will wash alkali down to the lower, heavier lands and damage them. In order to prevent this the canal should be protected &om leakage and great care exercised in the use of water on SUMMARY. 119 the light soils of the upper slopes. Wherever the level of standing water begins to rise and threatens to come near the surface the land should be underdrained. Protection from seepage is sometimes afforded by a seepage ditch, providing the ditch is dug along an impervious layer which naturally rises, and therefore raises the seepage waters. If no such layer is present a seepage ditch would afford little protection. Where irrigation waters carry a high percentage of salts, or where the soil is naturally salty, surface flooding is thought to be better than furrow irrigation. By furrow irrigation a crust of salt accumulates along the tops of the furrows to such an extent that in time it may prevent the growth of plants, except in the bottom of the furrows where the water washes down the accumulation of salt from the previous irrigation. By flooding, the salts over the entire surface are washed down and, if drainage be good, they will pass off through the under- ground drainage, but if the drainage be poor they will return to the surface. To prevent as far as possible the return of salts to the surface, the growth of shading crops or thorough surface cultivation should be resorted to. There are some plants more resistant to alkali than others, and by the cultivation of these, alkali lands may be at times utilized. Such plants as the Australian saltbush, sugar beets, sorghum, and sweet clover are worthy of trial. It is found that sweet clover will grow successfully with as much as 1 per cent of salt in the soil. It is a biennial, and therefore easily removed from the land. It is a deep-rooted legume, and consequently loosens the soil and supplies nitrogen. Its blossoms furnish excellent bee food and the leaves and stems are eaten by cattle to some extent. It serves the twofold purpose of improving the land and furnishing forage. SUMMARY. The discussion of the conditions in the three counties included in the reconnoissauce shows, in all the cases cited, that the presence of alkali salts within the soils is due to defective natural drainage. The obvious remedy is artificial drainage. Where farming is as intensive as it is in some parts of Utah the flrst step in the natural reclamation of the land, after application of water, should be drainage. This feature of the irri- gation question has been neglected and the result has been the damag- ing of valuable land. The method which should now be adopted for the reclamation of the lands abandoned on account of seepage waters and alkali is the same method which, if originally applied, would have insured against such damage— that is, underdraiuage. Tile drains offer the most effective and, in the end, the cheapest method of draining the soil. The application of gypsum or land plaster can be recommended for the black alkali soils. The effect of the gypsum on such soils is two- fold : First, the gypsum, improves the physical condition of these heavy lands, that is to say, it renders the soil more loamy in character and 120 REC0NN0IS8ANCE IN UTAH. easier to cultivate and drain; and, second, it neutralizes the black alkali, turning it into the white salt. Gypsum is not a chemical anti- dote for the white alkali, and its application to white alkali lands can be of little value, except in so far as the gypsum improves the physical properties of the soil. The growth of deep-rooted crops tends to loosen the soil and to per- mit the free downward movement of the water. This effect, together with the shading of the ground in order to prevent evaporation and the consequent formation of alkali, is well illustrated in the growth of alfalfa. A RECONNOISSANCE IN THE CACHE A LA POUDRE VALLEY, COLORADO. By THOS. H. MEANS. SOILS AND ALKALI. The Cache a la Poudre Valley, or the Greeley country, as it is some- times called, is one of the oldest as it is one of the most prosperous irrigation districts of the West. One month was spent in a reconnoissance of this valley in the sum- mer of 1899, and as the conditions there are unique in a way the results of the investigations may be of more than mere local interest. The Cache a la Poudre Valley, or at least that part of it which lies east of the foothills of the Eocky Mountains, is cut from a series of nearly horizontal stxata of cretaceous rocks. That portion of these strata which is exposed in the irrigated parts of the valley is largely sandstone or sandy shale, though at some places a bed of heavy blue shale is exposed. Prom these shales and sandstones the soils of the valley are formed, modified in a measure by the mixture of materials brought down from the mountains by the streams. The farming lands are situated on a more or less perfect system of terraces extending back from the stream and merging indefinitely into the upland, which con- sists of hills rounded by erosion. The lateral valleys, which extend back from the main Poudre Valley, are the result of flood erosion, no water flowing through any of them before irrigation was practiced except during times of heavy rains. The soils of the bottom lands are generally sandy or gravelly, with coarse, gravelly subsoils. Such soils, where well drained, furnish excellent truck lands 5 but the greater part of this low land is wet from the seepage from the upper irrigated lands, and at present is used for pasturage and hay crops. The second and third bottoms consist of heavier soils, becoming in some places a heavy clay. This clay land, on account of its imper- vious nature, acts as an obstruction to the flow of seepage waters from the uplands, and is in consequence often wet or alkaline. This type of land abuts directly against the upland soils, and extends around the mouths of the draws and lateral valleys. The same type 121 122 RECONNOISSANCE IN COLORADO. of heavy soil is found in the bottoms of the draws, where the natural moisture has collected the fine particles of soil from the hillsides and promoted the disintegration of the large grains. The greater part of the fanning lands of the valley lies upon the rounded hills of the uplands. The soil ou these hills is a sandy loam in nearly all cases, varying slightly in texture. Immediately north of Greeley it contains about 7 per cent of clay, while north of New Wind- sor it contains 12 per cent, and around Fort Collins the soils are as a rule still heavier. Many inclosed basins are found throughout the country, in the bot- toms of which the soil is heavy and impervious. In these natural basins the water collects after each rain, and when they are irrigated the waste water collects in the lowest parts, forming swamps. The bottom lands were the first to be irrigated, and as the country became more thickly settled new canals were built covering the higher lands. In this way the irrigated land has extended back from the river to a distance in some places of more than 10 miles. The construction of one canal above another in this way has opened a wide field for inquiry into the possibilities of damage from seepage water, and in the present investigation special attention was given to the damage already done and to the possible remedies for this damage. Since the time allotted to the field work was too short to warrant a complete study of the district, two townships were selected, comprising a strip of land from the Poudre to the desert land above the uppermost canal, the Larimer County Canal. This section was studied in detail, and the areas of wet or alkali soil were outlined In the field on a map. In this way a definite idea was obtained of the amount of wet land which at present exists, and the best means for removing this excess of water were considered. It was found that the amount of wet land under the Larimer County Canal is small and confined to the bottoms of draws and. laud imme- diately adjoining the canal. The whole of the area under the Larimer County Canal is not farmed at present, but when all the land is farmed this area of wet land will be likely to increase. The seepage water, however, seems to originate in the losses from the canals and constantly running laterals rather than in seepage from the irrigated fields. Professor Carpenter, of the Colorado Experiment Station, has investi- gated the question of the origin of the seepage waters, and, in his opinion, the greater part comes from the canals and laterals. The material through which the canals run is largely loose iu char- acter, and the water is clear, carrying very little material which would clog the interstitial spaces of the soil. There can be no question, how- ever, but that over irrigation, through ignorance or neglect, is also the cause of much seepage. The effect of over irrigation is very noticeable in some districts, and the careless use of water can not be too strongly condemned. It not only injures the land to which it is applied, but it SOILS AND ALKALI. 123 also largely increases the seepage water to the destruction of lower lauds. Under the Larimer and Weld Oaual the amount of actual damage is greater than under the Larimer County Canal. The lands have been under irrigation longer and the subsoil has had more opportunity to fill up; besides, the amount of land irrigated is larger. This laud also receives the seepage from the lands under the Larimer County Canal. Under No. 2 canal the amount of wet land is still larger than under either of the canals above mentioned. It receives the seepage from all the land above, as well as the seepage from several reservoirs situated at a higher elevation. The soils under 'No. 2 canal are heavier than the upland soils, the land is more level, and the natural drainage is poorer. In the district mapped several large areas of wet land under this canal are shown. This area receives the direct seepage from all the upland, and since the underlying beds of gravel are not contiguous to give ade- quate drainage, the water tends to rise to the surface in places and swamps are formed. One of tlie first questions which should be considered in the opening of new farming lands is the drainage. If the natural drainage is good — that is, if the excess of water is quickly i emoved from the subsoil — the installation of drains is not necessary, but if the water at any time stands within the subsoil or if the excess of water, applied through irri- gation or falling as rain, does not quickly pass away through the under- ground drainage the crops grown upon the land will suffer and the farmer will not obtain the best results from his efforts. The wet laud of an irrigated country should be immediately drained when the level of water rises closer than within 3 feet of the surface of the ground. It may be that the water rises near the surface only during a limited period of the year, but this may be long enough to injure a crop. The underground waters of an irrigated district situated within the arid regions of the West are never free from salts in solution. When this water is allowed to approach the surface of the ground it evapo- rates, leaving its burden of salt on or near the surface. The salt con- tinues to accumulate in this manner unless the surface water is drained away. Usually such a quantity of salt accumulates that nothing use- ful will grow upon the land. When reclamation is attempted both the water and salt have to be removed, thus making the work of reclama- tion very difflcult and costly. Prom the standpoint of economy, there- fore, it is much better to install drains before or at least as soon as the ground becomes wet. This will not only remove the excess of water, but will insure the land against ever becoming alkaline. In the Greeley district the process has as yet gone in most places only far enough to damage the ground from water. If the water is not removed much more damage is probable from alkali. A wet piece of ground is valuable in some cases as pasture, but a piece of badly alka- line land is practically worthless. The underground waters are not 124 RECONNOISSANCE IN COLORADO. highly charged with salts and the evolution of an alkali flat is slow, but none the less sure. In the shales of the underlying rocks quantities of alkali are stored, and where the seepage water passes through this shale and appears again the accumulation of alkali at the surface is much more rapid. Repeated tests were made for sodium carbonate, but none was found. This was to be expected, since all of the soils contain small quantities of gypsum, which is the chemical antidote for black alkali. There is no chemical preparation known which would render the alkali of the Poudre Valley harmless, consequently in order to redeem the lands already damaged these salts must be removed from the soils and removed so far that there can be no possibility of their ever coming back again. There is but one known way of effectively removing these alkali salts, and that is by underdralnage. In the Poudre Valley at present only the lower lauds are in need of drainage. The lower lands along the Poudre Elver and the immediate bottoms of the draws extend- ing back into the hills should be drained at once. In some cases a sim- ple line of drains up the center of the draw would sufSce for the present and would insure much of the bottom land from damage. By drawing off the water from the hill land through proper drains in the bottom of the draws, much less water would reach the Poudre bottoms as seepage. The larger tracts of land in the draws and Poudre bottoms should be thoroughly tiled. The shallow basins and sinks offer the most serious difficulties in the construction of drains. Where the basin has a shal- low depth the expense of cutting an outlet for the water may be slight, but where the basin is deep the expense of cutting an outlet is liable to be great. SUMMARY. Considerable damage has been felt in parts of the Poudre Valley from wet or alkali soils. Such wet or alkali tracts are the natural result of poor drainage. Tile drains should be installed in all the lower lands both to remove the excess of water and to prevent the accumulation of alkali. With continued irrigation of the uplands the amount of possi- ble damage to the lower lands is very great and to insure against this damage the drainage should be commenced at once. SOIL SURVEY IN THE CONNECTICUT VALLEY. By CLARENCE W. DORSEY and J. A. BONSTEEL. INTRODUCTION. In the present appropriation bill for the United States Department of Agriculture, Congress has specifically authorized the mapping of the soUs of the principal tobacco areas of the United States. On the Ist of July, 1899, the Secretary of Agriculture authorized this work to be started in the Connecticut Valley. About three and a half months were spent in surveying the soils of a portion of the valley. The area is comprised within north latitude 41° 40' and 42° 17', and west longitude 72o 30' and 72° 45'. It extends from South Glastonbury, Conn., where the valley pinches together, northward for a distance of about 41 miles, to Bachelor Brook, in South Hadley, Mass., where the Mount Holyoke range of mountains completely separates it from the extension of the valley from Northhampton northward into Vermont. Thevalley has an average width of from 5tol0 miles on either sideof the Connecticut River. The area surveyed and mapped comprised approxi- mately 400 square miles, or 256,000 acres. The object of the work was primarily to investigate and map the dif- ferent tobacco soils, but incidentally all soil areas were surveyed. Eather full notes were taken as to the general condition of agricultural practice in the valley, the condition of labor, the improved implements used, the construction of barns and other farm buildings, transportation, and other matters contributing to the agricultural features of the locality. Particular attention was paid to the kind of tobacco and the influence of the different soils upon the texture and quality of the tobacco. As this soil work, however, is but the basis of a very extensive and sys- tematic investigation into the physiology of the tobacco, and into the possibilities of changing the type and character of the tobacco through cultural methods and fermentation by the tobacco expert of the division, many of these notes will be reserved for future publication by the division, enough only being given here to make the soil work intelligible and interesting. One fieature which has been very clearly recognized in the course of the survey is the continual and rapid encroachment of city and subur- ban development for summer residences and for industrial purposes, 125 126 SOIL SURVEY IN CONNECTICUT VALLEY. Many extensive areas which were formerly considered agricultural lands are now built up or held for speculative purposes for residence or industrial pursuits. TOPOGRAPHY OF THE VALLEY. The Connecticut Valley is bounded on either side by hills rising to an elevation of 50 to 100 feet above sea level in the neighborhood of Hartford, and to a little over 500 feet in the northern extension of the area. The difference in the elevation of the river from the upper to the lower portion of the area is but slight, and few falls or rapids occur within this distance. Some water power, however, is developed at a few places along the river. The country is level or gently rolling, slop- ing gradually back to the high rounded hills and low mountains which form the boundaries of the valley. At places there are still well-defined terraces with sharp escarpments with elevations varying from 10 to 100 feet above sea level. GEOLOGY. The origin of the soils is partly glacial, which is seen in the great drumlins, or hogbacks as they are called, and the heterogeneous mass of bowlders, sand, and clay bordering the valley and derived from all sorts of rocks; and partly from a shallow glacial lake which is sup- posed to have spread out over the valley from a dam somewhere below Hartford. Into this lake sediment was brought by rivers and streams. This sediment was sorted over and spread out more or less evenly over the bottom of the lake. As is usual in such cases the deeper and quieter portions of the lake received the finer sand and clays, while the coarse sand and gravels were deposited near the shore line and near the mouths of rivers. These deposits were evidently laid down during the glacial epoch, as arctic plants and leaves are occasionally found in the thin layers of clay and shale, indicating a very different climate than that which now prevails. After the lake was drained the Connecticut River and its tributaries commenced cutting a series of terraces through the valley. These ter- races are not very well preserved, and can not be followed for any great distance, but they are very plain in certain parts of the valley. The character of the formations laid down in these two ways will be described more at length in connection with the soils. CLIMATE. The temperature of the Connecticut Valley ranges from 56° F, in May, to 610 in September, with an average of 70° in July, which is the hottest month of the year. The mean maximum temperature ranges from 69° in May to 82° in July, with a mean dailj-^ range of 20°. There is on an average about 4.5 inches of rainfall during each month of this Report No, 64. U. S. Dept. Agr. PLATE XXI. TOBACCO. 127 growing season, while the mean relative humidity during June, July, and August is about 70 per cent. Comparing these conditions with the climatic conditions of the other tobacco districts of Cuba, Sumatra, Florida, and Pennsylvania, it would seem that the temperature is sufficiently high for the production of either a wrapper or filler leaf. The rainfall appears to be sufficient, provided it is well distributed during the season; but the daily range in temperature is much greater than occurs in either Cuba or Sumatra, and this very likely has a great deal to do with the character of the leaf. With so great a daily range it would tend to thicken the leaf and increase the body. This is counterbalanced, to a large extent, by the light, sandy character of the soil, which is naturally adapted to the pro- duction of a thin leaf, provided a rapid and uninterrupted growth can be maintained. TOBACCO. The most interesting and most prominent feature in the agriculture of the Connecticut Valley is the tobacco industry, which has given a world-wide reputation to the locality, and has provided work and sus- tenance to a large number of people. Tobacco was introduced into the Connecticut Valley as a recognized farm crop in the early part of this century. It was early recognized that it diifered greatly in its qualities from the Maryland and Virginia tobaccos. It had less nicotine, less body, and was not so well adapted to pipe smoking or to chewing. It is an interesting historical fact that the first cigars made in this country are reported to have been made in the Connecticut VaiUey about the year 1802. It may be said to have been the home of the domestic cigar tobacco, as it certainly was the home of the domestic cigar manufacturing. With the specialization which has since developed in all lines of tobacco industry the Connecticut tobacco has taken its place essen- tially as a wrapper leaf, and it is not used to any extent at the present time for fillers in domestic cigars. It is essentially a light wrapjier, and when dark, heavy wrappers were in style, as periodically happened, especially before the wide introduction and extensive use of the ideal Sumatra wrajjper, the cultivation of tobacco in these light soils of Con- necticut was largely abandoned and the domestic supply of wrapper leaf came from the heavier soils of Pennsylvania. The torn, coarse, or inferior leaves are used as binders, while the trash and waste from the barn and cutting tables are exported mainly to England and brings from 1 to 2 cents per pound. The characteristics of a good wrapper leaf, as described by Mr. Floyd,' are that it should have but little body, little aroma or flavor, should be very pliable so that it will stretch and cover well, and have good tex- ture, grain, and style, in order that it will appear well on the cigar. 1 Report No. 62, U. S. Department of Agriculture. 128 SOIL SURVEY IN CONNECTICUT VALLEY. The leaves must be of uniform color and not too large, the 14 to 16 inch leaves being the most desirable sizes. While the Connecticut tobacco lias long been recognized by the trade as the most desirable domestic tobacco for wrapper purposes, yet the difference in price shows at once how the tobacco is regarded by manu- facturers in comparison with the imported leaf. The Conuecticuttobacco is worth, on an average, about 18 or 20 cents per pound; the Sumatra tobacco, imported exclusively for wrapper purposes, pays a duty of $1.85 per pound and sells on the market for from $2.50 to $3 per pound, duty paid. The Connecticut leaf is too large for an ideal wrapper, being often from 26 to 30 inches in length, the veins are very large, and only the tip of the leaf is suitable for high-priced cigars. Either on account of the physiology of the leaf or in the method of case sweating the desirable grain, color, and style are confined to the tip of the leaf, the lower half being glossy and very undesirable for wrapper purposes. This makes a great deal of waste, which can only be marketed in foreign countries at an exceedingly low price. Lastly, the tobacco is more highly flavored than is desirable for wrapper purposes and frequently masks the desirable qualities of the filler used in the cigar. These defects, as already stated, are to be made the subject of an exhaustive inquiry in the Division of Soils. One of the objections urged by the manufacturers against the Con- necticut tobacco, a fact which certainly largely reduces the price i)ai(l for the crop, is the unevenness of color and the poor grading as to color, length, and quality of leaf. In order to maintain a uniform brand of cigars, a manufacturer is forced to purchase a large amount of Connect- icut leaf from which to select. Furthermore, on account of the dift'er- ence in length and in texture of the leaves, there is considerable waste, which is difWcult to estimate. It is hard to plan, therefore, for an eco- nomical use of the product when a purchase of this tobacco is made. This is not the case with the carefully sorted Sumatra. One of the reasons which makes the quality of the leaf uncertain and varied is believed to be the method of fermenting in cases, the result of which is largely uncertain and dependent upon chance conditions, which are difficult to understand and impossible to properly control. It is believed by the tobacco experts of the Department that the bulk method of fermentation, as practiced in Florida and in Cuba, will give much more uniform and more desirable results. The principal reason why the assortment and grading are not so closely done is the high cost of labor. There is so much demand for labor in the shops and factories that the farmer is forced to pay a high price for help. The most successful farmers, therefore, have been those who have cultivated small tracts of from 5 to 15 acres in extent and have made the crop by themselves, with such help as their families could supply, and with occasional hired labor. The Connecticut farmer has CO pay about $1.50 per day, where the Florida planter pays about SOILS. 129 50 cents per day for labor. It has always been found, however, that in the exhilarating climate of the North the more energetic laborers do so much more work in a given time and to so much better purposes that successful competition is possible iu inauy lines along which at first sight the outcome would appear at least uncertain. It must be remembered, furthermore, that much of the most successful tobacco growing in Connecticut is done by the farmer and his family, so that a better system of fermentation and of sorting and grading could be done by them during the slack time of the winter months, when other work was not pressing, and without greatly increasing the actual expendi- ture of money in the production of the crop. In Sumatra the cost of labor is very much less, even, than in Florida, and this is one reason why the Sumatra planters are able to give the extreme care to the assortment and classification of their tobacco. It must be remembered, also, that in a fancy assortment of well-cured and choice tobacco the price is largely speculative, as there is no sharp basis upon which to fix a commercial value. When from the superior excellence of the leaf the price goes beyond 30 cents per pound it is liable to increase in wider and wider units to $1, $2, and even |3 or $4 per pound, according to the fancy of the purchaser and the skill of the producer in working up a trade and supplying special demands. This has been shown in a very marked way in the development of the tobacco industry in Florida, where high prices are obtained by a few packers who understand the market requirements and can judiciously place their products, while a very low price is obtained by the average grower, who is less able or less willing to make the product required by the trade, and who is unable to make advantageous trade connections. There is one interesting feature in regard to the tobacco industry in the Connecticut Valley, namely, that through the improved methods of planting and cultivating the crop the season of growth is being very materially shortened. In 1899 the growing season was about two weeks shorter than was ever known before. Tobacco set out the middle or last of May was fully matured by the middle of July. SOILS. The object of the soil survey is to classify and map the soils accord- ing to any condition which might influence the character of the vegeta- tion, especially the character of the tobacco, the kind of crops adapted to the land, and the quality as well as the quantity of the crops grown. The soils of the Connecticut Valley have been classified in this way as a result of the season's work, and a map has been prepared, accom- panying this report, showing the area and distribution of the various types of soils which were recognized. As before stated, this work will form the basis for a more extensive investigation of the Connecticut tobacco, and, pending such fuller investigation in regard to the char- acter of the tobacco, only brief mention will be made at this time of the H, Doc. 399 9 130 SOIL sriiVEY IN CONNECTICUT VALLEY. relation of tlic soils 1o the character of the leaf wbicli they produce. So many otbcr (inestions enter in regarding the variety of seed used, the method of cultivation, fertilization, and fermentation, as well as the personal equations of the growers, that require the careful investiga- tion and Judgmeiit of the tobacco expert, that only passing mention should be made at this time and in this connection of the character of the crop produced on the different soil areas. TKIASSIC STONY LOAM. As already mentioned, the Connecticut Yallev is bounded on either side, from the southernmost part of the area surveyed up to Westfield and Mill Eiver, just below Springfield, by a glacial deposit consisting mainly of Triassic sandstone debris forming long lenticular hills, drumlins, or hogbacks, between which the surface is rolling and hilly, with a few long ridges and groups of rounded hills varying in elevation from 40 to 400 feet. A few areas of this glacial debris are scattered about in the valley proper, and from the fact that they are not covered with lake sediment it may be assumed that they were formerly islands in the glacial lake, or else the covering has since been removed by erosion. The Triassic stony loam soils are fine sandy loams, dark Indian red in color, mixed with gravel and bowlders of all sizes and shapes, varying in size from an inch to C or 8 feet in diameter. The amount of gravel and undecomposed rock exceeds 5 per cent in all cases, and may exceed 50 per cent. In many cases fields are now comparatively free from these bowlders and large rocks on account of the numerous times they have been picked over and stones removed. Indeed, the surface of this geological area has been very greatly modified by the hand of man. Cultivated fields may contain only a few scattered stones, while the surface of the area surrounding the field may be covered with a mass of stones and bowlders. The stones, especially the smaller ones, and the fine gravel are derived principally from the Triassic sandstone and shales. Tobacco is grown to a considerable extent in certain areas in the Triassic stony loams. It is considered a fat tobacco, and has an unde- sirable cinnamon color, but it is readily bought up at good prices for special market demands. The character of the tobacco is quite different from that grown on the other tobacco soils of the valley, and there is a longstanding controversy as to the relative merits of this and of the tobacco from other soils. The fact of the matter is that they are used largely in different channels, and are to a great extent bought up by different dealers. There are large tobacco centers around Day Hill, south of Poquonock, around Warehouse Point, Suffield, and Enfield street. The best development of the tobacco industry on this formation is probably around Suffield and Warehouse Point. The Havana seed is exclusively grown on this formation. The other areas outlined on the map have soils of fair condition, but are not farmed to any great Report No. 64, U. S. Dept. Agr. Plate XXII. Report No. 64, U. S. Dept Agr. Plate XXIII. HOLYOKE STONY LOAMS. 131 extent. They are given up mainly to pasture lands, meadows, and orchards. The following table gives the mechanical analysis of the Triassic stoiiy loam subsoils: Mechanical analyses of subsoils of Triassic stony loam. (Fine earth.) Diameter. Millimeters. 2 to 1 ] to 0. 5 5 to .25 .25 to .1 .1 to .05 .05 to .01 .01 to .005 .005 to .0001 Conventional names. Gravel Coarse pand- ... Medium eaiid.. Fine sand Very fine sand. Silt Fine silt Clay Loss at 110° C . . . Loss ou ignition. 4211. Bloomfield, ^ mile S. Per cent. 2 3.35 8.60 31.25 34.22 4.35 6.20 6.57 1.36 2.03 4212. Enfield. Per cent. 12.45 11.86 13.98 14.78 17.51 8.20 8.67 10.23 1.04 1.09 4201. Hazard- ville, IJ miles S. Per cent. 5.26 8.G6 18.83 21.00 18.83 8.70 5.30 10.87 1 01 1.77 HOLYOKE STON'Y LOAM. The hills bounding the northern extension of the area examined, with elevations from 240 to 500 feet above sea level, are likewise glacial deposits, but formed of diabase, crystalline, and metamorphosed rocks. The surface of the country of what may be called the foothills is rolling with steep slopes and containing many lenticular hills and groups of rounded hills. The surface is rough, and in places large masses of bowlders are lying as they were piled by the ice, and with hardly any perceptible disintegration. Many of the slopes are entirely covered with these bowlders. The soil is a sandy loam, containing from 10 to 50 per cent of gravel and bowlders, ranging in size from 1 inch to 12 or 15 feet in diameter. The following table gives the mechanical analysis of the line earth of a sample of Holyoke stony loam: Mechanical analysis of subsoil of Holyolce stony loam. Millimeters. 2 1 0.5 .25 .1 .05 .01 to 1 to 0.5 .005 to .26 .1 .05 .01 .005 .0001 Conventional names. Gravel Coarse sand — Medium sand.. Fine sand Very nne sand. Silt Fine silt Clay Loss at 110° C... Loss on ignition . 4203. Aslileyville, 2 miles S. Per cent. 3.05 3.85 8.22 11.53 29.82 21.28 6.45 12. 20 1.54 2.35 The soils are derived entirely from the glacial debris. The soils are not fertile, and are not farmed to any great extent. There are occasional patches of corn, oats, and rye, but no tobacco. They are mainly given up to stony pastures and orchards. 132 SOIL SURVEY IN CONNECTICUT VALLEY. WINDSOR SAND. The generally accepted idea of the origin of the soils of the valley proper has already been referred to. It is supposed that during glacial times a dam extended across the valley below Hartford, forming a shallow lake of considerable extent, and comprising all of the area which has been surveyed this season. The streams collecting the drain- age from the surrounding country emptying into the lake brought sediments, which were sorted over and deposited in different places, according to the direction of the currents, the depth of the lake, and the velocity of the water. The coarser sands and gravels would be deposited near the shore line and the mouth of the streams, while the finer sands and clays would be deposited off shore, and in deeper and quieter i^ortions of the lake. As the level of the water of the lake was lowered from time to time well-marked shore lines, constituting terraces, were formed out of the material which had previously been deposited, and out of fresh material which had been emptied into the lake. This gave rise to a series of terraces, more or less well preserved at the present time, but greatly cut up and modified by recent stream action, and by the meanderiugs of the Connecticut Eiver. The Windsor sand represents what is supposed to be the original bottom of the old glacial lake in its shallowest parts. The soil is com- posed of yellowish-red or brown sand, resembling a coarse sharp build- ing sand, and containing less than 5 per cent of clay. The material contains about 5 per cent of coarse gravel, ranging in size from 2 to 10 mm. in diameter. This gravel content increases to some extent in the subsoil. In places this formation is 40 feet deep. The following table gives the mechanical analyses of four samples of Windsor sand : Mechanical analyses of subsoils of Windsor sand. Millimeters. 2 to 1 1 to 0. 5 0. 5 to . 25 .25 to .1 to . On to .01 to . 005 to 1 05 01 005 0001 Conventional names. Gravel Coarse sand Medium sand.. JTine sand Very fine sand. Silt' Fine silt Clay Lo.ssatllOo C ... Loss on ignition . 4210. Bloomfield 2 miles SE. Per cent. 4.98 11.31 33.41 33.75 10.82 2.09 1.03 1.65 .50 .80 4198. Windsor Loclcs, Smiles SW. Per cent. 8.00 16.83 39.90 22.21 6.66 3.20 1.20 2.25 4199. Hubbards Corners, IJ miles If W. Per cent. 2.52 12. 32 39.29 27.92 7.30 4.33 1.62 3.15 .48 1.84 4200. Chicopoe, 2 miles SW. Per cent. 14.30 26.28 30.35 9.55 6.03 4.66 2.73 4.87 .60 1.47 The soils are coarse and inclined to be leachy or droughty, but, like many soils of this character, are generally somewhat moist beneath the top few inches of dry sand throughout the most severe drought. The soils are good representative truck soils and are used for this purpose in some parts of the area. Report No. 64, U. S. Dept. Agr. Plate XXIV. a O O o o o HARTFORD SANDY LOAM. 133 In favorable seasons a very fine quality of thin-leaved silky tobacco is produced on these soils, the finest probably that is produced on any of the soils of the valley, but the season has to be just right, and such favorable conditions hardly occur more than two years out of five. If a crop by any means can be kept growing rapidly and continuously, it will produce a fine silky leaf excelled by none of the other soils in the valley. In average seasons, however, the growth is liable to be checked by changing weather conditions, and the plant is small and produces a leaf which is thick and strong. It is desirable to plant early on these lands and give constant and thorough cultivation. The Windsor sand comprises a large part of the Windsor Plains in the towns of Windsor, Bloomfleld, East Granby, and Windsor Locks, as well as a small area in East Hartford. The surface of the area in Con- necticut is level or gently rolling, ranging from 100 to 180 feet in eleva- tion. The area in Massachusetts is much larger and covers in all about 30 square miles. It is largely developed in the towns of Agawam, West Springfield, Springfield, Chicopee, and Long Meadow. The sur- face is much more rolling than in Connecticut, and ranges in elevation from 100 to 240 feet. For the most part the area of Windsor sand is not very extensively cultivated at the present time. In Agawam there is some truck farm- ing, a few peach orchards, and some tobacco grown. East of Spring- field the area is rapidly building up in the extension of the city. In Connecticut there are but few houses, deep sandy roads, and many old and unsuccessful fruit farms. Many areas which were formerly culti- vated are now grown up in the characteristic forest growth of pine. The soil is so open and porous and offers so little resistance to the entrance of rain water that the surface hardly washes at all, and there are old corn rows running through the forests and well preserved, upon which the trees must be at least 50 to 80 years old. Many of the old furrows thrown up in the original measurement of the land are still plainly distinguishable. HARTFORD SANDY LOAM. The Hartford sandy loam occupies by far the largest extent, and plays the most important part in the tobacco industry in the Connecticut valley. It extends from Glastonbury to South Hadley, and covers in all an area of over 80 square miles. A portion of it is probably the undisturbed old lake bottom, but there are also more recent river-cut terraces. The formation occurs in broad terraces, which are very level in places and gently rolling in others. The formation is found at ele- vations ranging from 30 to 260 feet above sea level. The soils are red, brown or yellow, medium grade sandy loams, about 12 inches deep, underlaid with yellow sands, containing little or no organic matter. The soil is a grade finer than the Windsor sand, as is apparent from the analysis, and is correspondingly stronger, and decidedly safer as an agricultural soil. The deposit varies in depth from about 3 feet to 20 134 SOIL SURVEY IN CONNECTICUT VALLEY. feet. It diii'ers again from the Windsor sand in having no gravel, or but a mere trace of very fine gravel in a few places. The following table gives the mechanical analyses of five samples of Hartford sandy loam: Mechanical analyses of subsoils of Hartford sandy loam. Diameter. 2IilUmeters. 2 tol 1 too 5 5 to 25 25 to 1 1 to 05 OS to 01 01 to 005 005 to 0001 Conventional names. G-ravel Coarse sand . - - Medium sand.. Fine sand Very fine sand. Silt' Fine silt Clay 4204. Windsor, -SW. 'er cent. 2.20 7.51 33.60 32.05 13. 50 4.47 1.75 2.78 Loss at 110° C... Loss on ignition . .80 1.30 4205. East Long I\leadow3, 2 miles NW. Per cent. 0.00 .31 2.84 63.10 29.15 1.15 .96 1.42 .60 .90 4314. Burnside, J mile SW, Per cent. 2.23 7.73 26.25 20. 00 25.40 3.45 2.10 3.22 .77 1.27 4213. Sonth "Windsor, 2 miles £. Per cent. Trace. 6.84 42.86 33.00 7.73 2.63 1.70 3.50 .75 1.54 4215. Burnham, 1 mile E. Per cent, 4.11 11.83 29.20 24.45 12.72 3.48 3.28 5.20 2.95 2.81 The general crop of Connecticut seed-leaf tobacco is grown on these soils. This represents what may be called the typical tobacco soil of the Connecticut Valley, and the safest and therefore the' best soil at least for the seed-leaf variety. Corn and potatoes do well on these soils, and about Hockanum and south of Willow Brook, in East Hartford, con- siderable truck farming is carried on. Southwest of Hazardville the soil for some reason appears less coherent, and drifting sand dunes are common. The yield of tobacco on this soil varies from 1,500 to 2,000 pounds per acre, or even a little more. The average price of the crop is about 18 cents a pound, but the lands are heavily fertilized. Labor is high and the crop as a whole is expensive to make. Tobacco is grown mainly in small fields of from 3 to 10 acres, 15 acres being an unusually large tract. There are many large areas of this still unde- veloped upon which a good quality of tobacco can be produced: PODUNK FINE SANDY LOAM. The Podunk fine sandy loam represents river cut terraces, ranging in elevation from 20 to 80 feet above sea level. There is in all an area of only about 7 or 8 square miles of this in the area surveyed, and of this only about J square mile in Bast Hartford has contributed largely to the fame of the Connecticut Valley in the production of the broad-leaf variety, which differs in many essential characteristics from the Habana seed-leaf grown on the Hartford sandy loam. The surface of the areas are level or very gently rolling. The soil is a dark-brown sandy loam, about 12 inches deep, and of most excellent tilth. The subsoil is a dark- brown sandy loam of the same texture as the soil, but differing in color as it has less organic matter. The following table gives the mechanical analyses of two samples of Podunk tine sandy loam: Report No. 64, U. S. Dept Agr. Plate XXV. Report No. 64, U. S. Dept, Agr. Plate XXVI. CONNECTICUT MEADOWS. Mechanioal analyses of subsoils of Fodiink fine sandy loam. 135 Diameter. Conventional names. 4206. Agawam, ImileNW. 4216. South Windsor. i mile NE. yiillimeteni. 2 to I 1 to 0. 5 0.5 to .25 .25 to .1 .1 to .05 .05 to .01 .01 to .005 .005 to .0001 Per cent. 0.00 .07 1.53 41.80 49.00 3.43 1.02 1.70 Per cent. 0.60 1.51 7.90 23.27 41.82 9.15 6.32 4.40 Coarse sand . Medium sand Very iine sand Silt Fine silt Clay Los8 at 110 ° c .60 1.08 1.92 3.68 The soils are a grade finer in texture than the Hartford sandy loams, and while well drained are strong, safe, and productive. The areas are all thickly settled, and, with the exception of that in the town of Agawam, are very highly cultivated and well cared for. The South Windsor and Kaubuc districts, j)articularly, are famous broad-leaf tobacco areas. The broad-leaf variety is heavier and has a thicker leaf than the seed-leaf, is generally darker in color, and is a better leaf for cutting purposes. It has not the undesirable "seed" flavor of the Habana seed-leaf, and is preferred by many smokers. It yields more per acre than the Habana seed on the Hartford sandy loams, but, with all, it has a rougher look in the cigar. For this reason it is not in such good favor with many cigar manufacturers as the finer, silkier Habana seed grown on the Hartford loams. It is considered better by the farmers in this Podunk region to plant tobacco continuously rather than to use a rotation. There are fields which have been continuously in tobacco for twenty-five years, and which it is claimed are as productive and pro- duce as fine tobacco as at any period of their cultivation. This broad- leaf variety is grown to a limited extent upon the small area of Hart- ford sandy loam east of Connecticut Eiver, near South Windsor, but, with this exception, it is confined to the Podunk soil. Corn and potatoes do well on this soil. The Agawam area has not been very successfully farmed for the past few years, for some unexplained rea- son, and there is a chance of development there which should certainly arrest the attention of tobacco growers. CONNECTICUT MEADOWS. The Oonnecticnt meadows occur as narrow strips on either side of the Connecticut Eiver from Holyoke south to Long Meadow, Mass., and irom Warehouse Point to South Glastonbury, Conn. There are smaller areas along the various tributaries of the Connecticut Eiver. This represents the present flood plain of the Connecticut Eiver and its tributaries, being built up at the present time by deposits from the flood 136 SOIL SURVEY IN C0\NECTI(3UT VALLKT. waters. The river will finally build up this terrace so high that it can no longer spread over it, and new terraces will be constructed at lower levels. The surface of the meadows is generally higher along the river bank and, of course, better drained there. There is a gentle slope down to the scarp connected with the upland or next higher terrace, and this area immediately below the upland is frequently wet and swampy. The meadows are all subject to overflow at time of very high flood, but in spite of this there is considerable farming on them, although at some risk of the loss of the crop. Surface of the meadows is level or gently rolling, with occasional old stream channels or run-ways in them. The surface of tbe meadow is from 5 to 20 feet above sea level at Hartford and from 40 to 80 feet above sea level at Springfield and Holyoke. The material of which the soils are composed has been brought in by the rivers at flood time, but tbere is no evidence of stratification, probably on account of the wind and rain action between the comparatively long periods between the floods. The character of the material is very uniform throughout all the areas. It is very fine sand and silt, being uniform in places to a depth of 10 feet and over, resembling loess in many of its characteristics. It is a grade finer than the Podunk fine sandy loam. The soil is from 16 to 18 inches deep, contains a large amount of organic matter, consists mainly of very fine grades of sand and silt with but little clay, and is considered extremely fertile and pro- ductive. The subsoil is grayish in color, but otherwise hardly different from the overlying soil. The following table gives the mechanical analyses of four samples of subsoil from the Connecticut meadows: Mechanical analyses of suhsoils of Connecticut meadows. Millimeters. 2 to 1 1 to 0. 5 0.5 to .23 to .1 to .05 to .1 to .005 to 1 05 01 005 0001 Conventional names. Gravel Coarse aand Medium sand.. Fine sand Very fine sand . Silt Fine silt Clay Loss at 110° C... Loss in ignition. 4,207. Cbicopee, 2i miles N. Per cent. 0.00 Trace. 0.37 6,87 50.86 28.49 6.60 3.35 1.10 2.60 4217. Hartford, % mile SE. Per cent. 0.00 Trace. 0.21 1,50 19.55 33.67 28.54 9.50 2.60 4.75 4218. S. Wind- sor, h mile NW, Per cent. 0,00 Trace, 0,36 2 14,78 36,50 27.17 13,40 2,50 3,54 4219, Windsor, ^mileSE, Pe, • cent. U,00 ,00 ,31 1,77 9,79 30.25 29.47 19.11 4.13 5,31 Generally the meadow soils produce a dark, heavy, low grade of tobacco. When dark cigars were in style, as they were fifteen or twenty years ago, these meadow lands produced a tobacco which was much more in favor than it is at the present time. Yery good tobacco can be grown on high meadow land if it be well drained, but there is an almost universal prejudice against it and it brings a low price from the mere association with the name of meadow land. There is an area of high meadow laud west of Bast Windsor hill which is never overflowed Report No. 64, U. S. Dept. Agr. PLATE XXVII. ENFIELD SANDY LOAM. 137 by the river, resembling tbe Podunk fine sandy loam and classed with it, upon -which there is a very good quality of tobacco produced, but which suffers from the mere association with the name of meadow land. A large proportion of the Connecticut meadows is used only for grass, several crops being cut in the course of the year. Areas where very well drained are used for corn and potatoes, and in some cases for celery and general truck farming. Below Merrick the meadows are diked to keep out the high water and insure the land from overflow, but this is the only place where this has been done to any extent. The Connecticut meadows, Podunk fine sandy loam, Hartford sandy loam, and Windsor sand represent four important and representa- tive soils of higher and higher average elevation and of coarser and coarser material, which represent the most extensive and most impor- tant soil areas in the valley. The difference in texture of these soils is very marked and very apparent to the eye, and this difference in texture determines to a large extent the relation of the soils to crops, and particularly to the quality of the tobacco produced. ENFIKLD SANDY LOAM. On either side of the valley, in terraces around the hills of the Trias- sic stony loams and filling up depressions at elevations ranging from 80 to 240 feet, there is a fine sandy deposit, resembling the Hartford sandy loam, directly over the Triassic stony loams which occurs at a depth of about 2 feet below the surface. The outward extension of the terraces grades into the Hartford sandy loam, where the Triassic stony loam is 3 feet or more below the surface. The underlying Triassic stony loam provides a retentive subsoil, and these soils retain moisture better and are rather stronger than the Hartford sandy loams, which they resemble in other respects. The following table gives the mechan- ical analysis of a sample of Enfield sandy loam : Mechanical analysis of siibsoil of Enfield sandy loam. Diameter. Millimeters. 2 to 1 to 0.5 to .25 to to to to 1 0.5 .25 .1 .05 .01 .005 to .1 .05 .01 .005 .0001 Conventional names. Gravel Coarse sand — Medium sand. - Finesand Very fine sand. Silt Fine silt Clay Loss atliooc Loss on ignition.. 4202. East Hart- ford, E. Per cent. 4.22 5.05 7.75 10.43 43.60 13.53 4.86 5.57 1.82 3.22 138 SOIL SURVEY IN CONNECTICUT VALLEY. A very good quality of Habana seed-leaf tobacco is grown to a limited extent on tliese soils, especially around Melrose, Osborne, Wap- ping, and soutli of Poqnonock. The cliaracteristic feature of this soil as distinguished from the Hartford sandy loam is the occurrence of the Triassic stony loam at a depth of from 12 to 24 inches. The areas along the eastern part of the valley are generally thickly settled. In other parts of tbe valley there are large areas still uncleared and uncul- tivated. 8UFFIELD CLAY. The SufBeld clays occur principally at Windsor, Bast Windsor, and SuiBeld, with occasional small areas along the scarp between the Con- necticut meadows and the upland. These soils are found at elevations varying from 40 to 220 feet above tide. The surface is gently rolling, except along the scarps, where the sloiDes are liable to be quite steep. The soils are heavy clay loams, grading down into drab clays with exceedingly poor drainage, making the surface cold and wet. The fol- lowing table gives the mechanical analyses of two samples of Suffleld clay: Mechanical analyses of suhsoils of Suffield clay. Diameter. MilUmeiers. 2 1 0.5 .25 .1 .05 .01 to 1 to 0. to . to . to . to . to . 5 25 1 05 01 005 0001 Conventional names. Gravel Coarse sand . - . Medium sand.. Fine sand Very fine sand. Silt Tine silt Clay Loss at 110° C. . . Loss on ignition. 4208. Suffleld. Per cent. 0.00 .2!) .40 .73 5 32.57 29. 10 25.65 2.17 3.53 4209. Riverdale, W. Per cent. 0.00 .00 .00 .15 11.27 38.58 24 23. 50 1.10 L50 The Suffield clays have little value as agricultural lands, and are mostly given up to grass and pasture. There are large areas which are not cultivated at all, and growing up in a tangled mass of worthless underbrush. Around Springfield, Holyoke, and Hartford there are large brickyards making brick of this clay. No tobacco is grown in any part of this area, as the soils are entirely unsuited to this crop. BLMWOOD LOAM. The Elmwood loam occurs la large areas around Hartford, Windsor, and South Hadley. The soil consists of a deposit of flue yellow sandy loam, about 24 inches deep, overlying laminated drab clays, similar in all respects to the SufBeld clays. The clay is a deep lake deposit, while the overlying sand is a shallow lake or river deposit. The following table gives the mechanical analysis of a sample of Elmwood sandy loam: CONNECTICUT SWAMP. 139 Mechanical analyais of suhsoil of Elmwood sandy loam. Diiimeter. Oouventional names. 4220. Elmwood, 3 miles N. Millimeters. 2 to 1 1 to 0. 5 0.5 to .25 .25 to .1 .1 to . 05 .05 to .01 .01 to .005 . 005 to . 0001 Gravel Percent. 0.00 0.45 3.76 17.53 59.82 7.03 4.65 4.00 Veryfinosand Bilt Clay no n 1.40 1.90 The surface of the ground is rolling and hummocky. The soil is found at elevations varying from 20 to 80 feet above sea level at Hartford, and from 160 to 200 feet at South Hadley. Around Hartford the Elmwood loam gives fairly good grass and pasture lands, but is not cultivated to any extent except for small garden patches. North of Hartford it is neither cleared nor cultivated. The area around Hartford is rapidly being built up in the suburbs of the city. A few small areas have been cultivated in tobacco, above Windsor, but the soil is not adapted to tobacco cultivation. CONNECTICUT SWAMP. Considerable swamp land and wet meadows occur scattered over the entire valley at various elevations. They occur generally along the scarp between the Connecticut meadows and the upland and along a large number of small streams flowing into the Connecticut Eiver. Again in the hollows between hills and slopes, where there is poor drainage, these upland swamps are liable to occur from any cause where the drainage is poor and the soil almost impervious to water, with swampy, wet conditions as a characteristic feature. IsTo matter at what elevation they may be found these are all classed as Connecti- cut swamps. The swamps are all of fresh water and they are all at present too wet for cultivation. Some of the areas can be easily drained, while others immediately along the river could not be reclaimed except at great expense. Most of them along the river are wet throughout the year. Some of them are sufficiently dry during the summer so that the coarse, rank meadow grass produced may be mowed and saved for rough forage. The character of the soils of these areas has not been determined. They are very small in extent and of little agricultural importance. Many of them represent merely obstructed drainage areas which by underdraiuage could readily be reclaimed. 140 SOIL SURVEY IN CONNECTICUT VALLEY. COJIPAKISON OF THE TEXTURE OF THE SOILS. The following table gives a clear idea of the texture of the soils of the Connecticut Valley, and explains to a great extent the distribution of crops and the difference in the characteristics of the tobacco grown on the different soil areas : Windsor Plains, 65 per cent gravel to medium sand, 30 per cent fine and very fine sand. Triassic stony loam, 35 per cent gravel to medium sand, 40 per cent fine and very fine sand. Enfield sandy loam, 17 per cent gravel to medium sand, 55 per cent fine and very fine sand. Diabase stony loam, 15 per cent gravel to medium sand, 40 per cent fine and very fine sand. Hartford sandy loam, 65 per cent medium to fine sand, 90 per cent medium to very fine sand. Podunk loam, 40 per cent fine sand, 90 per cent fine and very fine sand. Connecticut meadows, 50 per cent very fine sand, 80 per cent very fine sand and silt. Suffield clay, 35 per cent silt, 50 per cent fine silt and clay, 25 per cent clay. APPLICATION OF THE THEORY OF SOLUTION TO THE STUDY OF SOILS. By FRANK K. CAMERON. INTRODUCTION. In the course of the work of this laboratory within the last year we have had occasion to examine chemically quite a number of soils and soil crusts from the arid regions. These soils have been of the class known as "alkali soils"; that is, they contain a relatively large per- centage of water-soluble salts. The origin and presence of these salts have been discussed in the publications of this Division and elsewhere, so that we may assume, for the purposes of this paper, that this part of the subjecthasbeen made suflSciently familiar to those who may be' interested. The salts which go to make up these crusts or water-soluble constitu- ents of the soils are, in general, sodium chlorid, magnesium chlorid, cal- cium chlorid, sodium sulphate, magnesium sulphate, calcium sulphate, and sodium carbonate with, occasionally, much smaller quantities of other salts which may be regarded as relatively unimportant. It is usual to find several of these salt components together in an alkali soil. When sodium carbonate is a component of the salts present then the salt mixture is known as "black alkali," because of the charring and caustic effect of the sodium carbonate on the organic matter present giving a characteristic black color to the soil. In contradistinction, when sodium carbonate is not a component, these salt mixtures are known as "white alkali," because, in general, they do not show this darkening effect but efiloresce on the surface as a white crust if present in sufflcieut amount. It is to be remembered, however, that local ter- minology is not always in full accord with this distinction ; for instance, in New Mexico, in the Pecos Valley, there are tracts of laud, quite dark in appearance, containing much water-soluble salts or alkali, which are locally called "black alkali" land, although examination fails to show the presence of any sodium carbonate. The chemical examination or analysis of these water-soluble salts in the soils yields results which, in many cases, seem very remarkable at first sight in the relatively large percentage of salts usually regarded as sparingly soluble — notably sulphates and carbonates of lime — and the question how they have been dissolved and transported in such large 141 142 APPLICATION OF THEORY OF SOLUTION. quantities is a natural one. When we attempt to dissolve these salts in water, in order to remove them from the soil, just such surprising results are obtained in our solutions in the relatively large quantities dissolved under certain conditions. In order to understand these phenomena and to make our analyses intelligible it will be desirable to have in mind the laws governing solutions which apply to such cases. NATUEE OF SOLUTIONS. Before discussing these laws it seems well to state more precisely what is understood by the term solution, and as we are concerned only with solutions of solids in liquids we shall not stop to make any modifi- cations which other solutions would necessitate. The most striking characteristic of a solution is its homogeneity j that is to say, after it has come to equilibrium every part of the solution is just like any other part in all its physical and chemical properties. Two important exceptions should be noted here. If the solution be one of considerable height or depth the gravitation force comes in evidence, the lower portions of the solution are more dense, the higher portions less so, and there is a corresponding variation in other properties. But even here it will be observed that there is no sudden break in the properties. They will vary regularly and continuously as we go from the top to the bottom. Again, it has been observed in the case of solu- tions in contact with solids (the walls of the containing vessel, for instance) that in portions very near the surface there is often a concen- tration of the dissolved material. This phenomenon has been called ad-sorptiou. The growth of crystals on the surface of solids intro- duced into the solution is probably due to this cause; the difi&culty of washing certain solutions through a filter paper, where often a portion of the dissolved substance will linger after the most persistent washing, is a well-known example of this phenomenon. Another important characteristic of a solution is that all its proper- ties vary continuously with the concentration. The law by which this variation takes place in any specific case may not be evident, and in general it is not, but it may be taken as a well-established principle that there is no sudden jump or change of direction in any physical property, as the concentration gradually changes. The recognition of this fact has come within comparatively recent years. It was long supposed that, while this fact was in general true, there were well- marked exceptions. For instance, it has been known since the time of Gay-Lussac that the solubility of Glauber's salt (the decahydrate of sodium sulphate) changed suddenly at about 33° if the temperature was gradually raised. So do all the other physical properties of the solu- tion; such as vapor pressure, refractive power, conductivity, etc. Within comparatively recent years the explanation was offered that below 33° the solution was one of the decahydrate in water, and above that temperature the solution was something different, containing a LAWS GOVERNING SOLUTES. 143 lesser hydrate or the anhydrous salt. Later study has shown that this view must be modified. There is probably no change in the substance Actually in the solution, for in that case we would expect some irregu- larities in the change of physical properties. The difference below and above this temperature point, 33°, is in the equilibrium between the dis- solved substance and the undissolved salt which may crystallize from the saturated solution. It is really the nature of the undissolved salt at the given temperature when in contact with the solution that is the controlling factor, and we have in fact two solubility curves in this particular case which meet in a common point near 33°. When the solid decahydrate is warmed to this temperature it partially melts into a saturated solution of the anhydrous salt, the excess of anhydrous salt separating as such. We are now familiar with quite a number of similar instances of this character, but in every case investigated a similar explanation has been demonstrated, and this characteristic of solutions has been thor- oughly established as a natural physical law. Indeed, these two prop- erties just described are incorporated in the best definition as yet pro- posed for a solution, i. e., a homogeneous mixture whose properties vary regularly with the concentration. LAWS GOVERNING SOLUTES. It has been found convenient in discussing solutions to introduce certain terms, and, as it will greatly simplify what is to follow, they are here defined. The substance in which the solution takes place is known as the solvent, and the substance which goes into solution in the solvent is called the solute. From the study of the phenomena presented by solutions it has been very clearly shown that the solute is in a condition strictly analogous to a gas — that is, its molecules are perfectly free to move about among themselves within the limits of the solution. Just as the movement of the molecules of a gas results in a pressure ou the walls of a containing vessel, so the movement of the molecules of a solute will result in a pressure on the walls of a containing vessel; only the medium in which they are free to move — the solvent — must extend beyond the walls of the vessel and be able to pass freely through them. Vessels with such walls have been realized ; they have been filled with a solution, immersed in the solvent, and the pressure on the walls of the molecules of the solute — trying to diffuse into the outside solvent — has been measured quantitatively. This pressure, which has been called the osmotic pres- sure to distinguish it from the gas pressure known as vapor pressure, has been found to be actually the same in amount as the vapor pres- sure would be at the particular temperature, if the substance could be a gas at that temperature. It has long been known that certain relationships hold between the volume, pressure, and temperature for all gases, irrespective of what 144 APPLICATtON OF THEORY OF SOLUTION. the particular nature of the gas may be. The explanation for this, which is universally accepted at the present time, is that in a gas the particles of which it is composed (the molecules, if the molecular con-' stitution of matter be accepted) are possessed of such a large degree of freedom that they are entirely able to move about among each other without this motion being impeded by anything in the nature of the molecule itself— in its shape, size, or other property. This view is supported, for instance, by the fact that a gas will diffuse or spread out into space indefinitely (practically) if not confined in a retaining vessel, which neither a liquid nor a solid can do without first passing into the gaseous state; and by the further fact that when the volume is forcibly made small or the gas put under great pressure, whereby the molecules are forced together so that presumably the individual motions are affected by the peculiar nature of the molecules, the gas obeys the usual laws governing the volume — pressure — temperature relations less and less accurately. It follows that these volume — pres- sure — temperature relations are dependent only upon the number of molecules involved. Just so a solute by virtue of the presence of the solvent is supposed to separate into molecules which can move so freely among one another as to be independent of any inherent characteristic of the molecule itself and, consequently, will move or diffuse throughout the solvent until the solution becomes homogeneous. Only, instead of diffusing indefinitely into space, the ultimate volume which the solute may occupy, independent of any containing vessel with semipermeable walls, is the volume of the solution. Therefore, it would be expected that these volume — pressure — temperature rel ation s which hold for gases would also hold for solutes, as it is only the number of molecules of the solute which are involved iu determining them and not anything in the nature of the molecule itself. Tbis has been found to be the case exactly, and all the laws governing these relations for gases are also laws for solutes if we substitute the term osmotic pressure for the vapor pressure. DISSOCIATION OP ELECTROLYTES. The solutions with which we most commonly meet, and those which will almost exclusively interest us, are those in which the solvent is water and the solute or solutes are salts ; that is, substances formed by the action of an acid on a base. These substances, when dissolved, show wide variations from the laws which have just been described. These variations are, perhaps, best studied by a consideration of the osmotic pressures they exert, which are found in all cases to be higher than was to be expected. Furthermore, these pressures vary with temperature or concentration in a way entirely different from solutions in general. It was first pointed out by Arrhenius that these variations are characteristic of solutions of electrolytes; that is, solutions in which the electric current passes. By making use of a hypothesis DISSOCIATION OF ELECTROLYTiES. 145 which had been advanced by Clausius the pheuomeua Avere readily explicable. Clausius was led to assume that when an electrolyte was dissolved in water it separated, partially at least, into components or, as we now say, dissociates. These dissociated parts are chemical equiv- alents, and in the act of dissociating develop equivalent amounts of electricity with different signs. Thus sodium chlotid would dissociate into sodium, bearing a certain amount of electricity with the positive sign; and a chemically equivalent amount of chlorine, bearing an equal amount of negative electricity. In the notation of chemistry this is expressed + - NaOl^lSTa+Cl In an exactly similar manner sodium sulphate would dissociate as thus indicated : + + - Na2S04^1Sa+]Sa+S04 the negative electricity on the group SO4, or sulphion exactly equaling the positive electricity on the two sodium equivalents. Borrowing an expression introduced by Faraday, these dissociated parts with their electric charges are called ions. When the solution con- taining ions is put in an electric circuit, the ions bearing the positive charges are attracted to the negative pole of the circuit and the ions bearing the negative charges toward the positive pole, in accordance with the well known law of electricity. Retaining sodium chlorid as an example, we may illustrate the phe- nomena by a mechanical analogy, thus: £ - - + + ^ o C1....C1 Cma ClNa Na....lsra ^ , g - - + + I: + .t 01.... 01 OlNa ClNa Na....Na g ~ l5 01 01 OlNa Oma Na---.Isra ^ The chlorine ions on arriving at the positive pole give up their elec- tric charges, thus becoming ordinary chlorine, which escapes as the well- known gas or reacts on the water of the solvent to form hydrochloric acid and oxygen, which escapes as a gas. In a similar manner the sodium ions discharge their electric burdens, thereby becoming" ordi- nary sodium, and immediately react on the water to form sodium hydrate and hydrogen. It is to be observed, however, that the chem-~ ical behavior of the ions is entirely different from what it becomes when their electric charges have been lost. The illustration just used indi- cates that there would be a gradual accumulation of hydrochloric acid at the positive pole and of sodium hydrate at the negative pole. This is what actually happens, as anyone can readily test by drawing off portions of the solution near the poles and analyzing them. H. Doc. 399 10 146 APPLICATION OF THEORY OF SOLUTION. Each of these ions in solution behaves to :ill intents and purposes as an individual luoleuule and exerts its influence as such in determin- ing the osmotic pressure in the solution. Without going into details, it may be said that from an estimation of the osmotic pressure it is possible to determine the number of ions in a given solution. From the conductivity it is also possible to deter- mine the number of ions; and the agreement by both of these methods is of such a character as to make this idea of electrolytic dissociation or ionization one of the best established working hypotheses science now possesses. REVERSIBLE REACTIONS AND THE MASS LAW. Suppose a solution of a salt, for example sodium cblorid, be brought in contact with another salt, such as potassium nitrate. There will then be in the solution not only the sodium chlorid and the potassium nitrate, but the ions formed from them, which may be represented thus: + + - NaCl, KNO:„ Na, K, 01, and NO3. + But the Na ions and NO3 ions can form sodium nitrate and will form it to a certain extent, so that we shall have this salt also in the solution. In the same way potassium chlorid will be formed and the final state of the solution will represent an equilibrium between the solvent water and the various salts and ions which can be formed. It is obvious that the same condition could be obtained by starting with sodium nitrate and potassium chlorid; and the equilibrium could be displaced one way or the other by varying any of the constituents. Such a reaction is called reversible. Again, sodium chlorid in water partially dissociates thus: + - NaCl f; Na, 01 But if the concentration be altered by the addition or evaporation of water, the equilibrium is displaced and either more sodium chlorid will dissociate or more will be formed from the already dissociated ions. In fact, practically all simple reactions so far studied, whether in solution or otherwise, have been found to be reversible. Suppose a reversible reaction between A and B with the formation of and D. It may be expressed thus : A+B50+D Let p, q, r, s represent the active masses or the number of molecules per unit volume acting, respectively. Let K be the rate of combination of unit masses of A and B to form and D, and Let K] be the rate of combination of unit masses of and D to form A and B. HETEROGENEOUS EQUILIBRIA. 147 The magnitude of the reaction A+B ->. is Kpq, for it is directly pro- portional to the rate of combination and to the active masses, and it will be Kirs for the reaction 0+D ^. Therefore when equilibrium is established Kpq = Kirs. (1) This is known as the law of "mass action," or of Guldberg and Waage. It is conventional to take as the unit of active mass in solu- tions one reacting weight per liter. Applying this formula to the case just cited, , - + NaOl S Na+01, we would have Kipq=K2r, where p represents the concentration of the ion Na, q that of the ion + 01, and r that of the undissociated salt NaOI. From which r Ki ' which is known as the dissociation or ionization constant, and is a con- stant for constant pressure and temperature, but is different for differ- ent salts. For those acids and bases and their salts which are usually charac- terized as strong K; is found to be large, while on the contrary it is always small for the so-called weak acids, bases, and their salts. Since it is obvious that it is the active masses of the ions which make an acid or base weak or strong, the magnitude of K; is generally taken as the measure of the relative " strength " of an acid or base at the present time. Still applying formula (2) to such a case as - + NaCl ^ Na -f 01 - + we see that for every JS'a ion there will be a 01 ion — that is, the concen- tration or active mass will be the same, and ^=Kor p2 = Kr r ^ therefore the concentration of the undissociated salt is proportional to the square of the concentration of the dissociated salt. From which it follows that the greater the dilution the greater the proportion of dis- sociated to undissociated salt. HETBROGBNEOXTS EQUILIBRIA. So far we have considered only cases of homogeneous equilibrium, where we had but one reversible reaction and complete solutionr But there are certain types of heterogeneous equilibrium which remain to be considered. 148 APPLICATION OF THEORY OF SOLUTION. Supi)ose a salt S iu contact with a solution saturated with respect to it and containing therefore S' of it. S' will be partially dissociated into A and B. Let c be the concentration of the undissolved salt, Ci the concentration of the dissolved but undissociated salt, and a and b the concentrations, respectively, for the ions. When equilibrium shall have been established between S ;t S' the equation expressing the fact will be Ci but c, the concentration of the undissolved substance, is the same thing as its specific gravity — that is, a constant K2; therefore From the dissolved salt we get S' ^ A + B and ab _ ^ (5) from which, substituting from (4:), ab = K3 K4 or K, (6) which is known as the solubility constant. If the equilibrium remain undisturbed, this equation (6) will hold, but if it be disturbed there will be two possibilities, either ab is greater than K (ab > K) or ab is less than K (ab < K). If ab > K, then the reaction will go A + B -^ S' and S^ ^ S, and there will be a precipitation of the salt. If ab < K, the reaction will go S ^ S^ and S^ -> A + B, and there will be a further solution of S. If we have two or more salts brought into solution in water, as was indicated above, there will be formed also all the possible combinations of which the dissociated ions are capable, as well as the ions them- selves, and of all these possible combinations that one will be the first to precipitate the product of whose ionic concentrations shall first exceed its solubility constant. Suppose, for example, that to a + solution containing sodium chlorid IsTaCl and its ions Na and CI we + - add potassium chlorid which will yield its ions K and CI, one of which is common to the salt already in solution. Then the jjroduct of the ionic concentrations of both salts is thereby increased. It might hap- pen that this product for the salt already in solution was equal to its solubility constant; the increase in the product brought about -by the addition of the salt with a common ion will indicate a state of instable equilibrium; there will be a "forcing back of the dissociation" with formation of an excess of undissociated salt and its precipitation from the solution. Thus lead chlorid can be easily precipitated from a solu- tion by the addition of potassium chlorid. By adding strong hydro- CARBONATES AND LIME IN GREAT SALT LAKE. 149 chloric acid solution to a moderately concentrated solution of barium cMorid one can precipitate the latter in beautiful crystalline scales. A very interesting case, illustrating these views, which is to be found in nature, has recently come under investigation in this laboratory. CARBONATES AND LIME IN GREAT SALT LAKE. The water of the Great Salt Lake in Utah is a strong solution of salts. The most conspicuous base present is sodium, though there are also potassium, magnesium, and a small amount of calcium. There is reason to believe that strontium would be found on a careful examina- tion, but it is present (if at all) in amounts which could easily escape detection without unusual refinement in the analysis. . Among the acids present carbonic acid has not been recorded in any analysis of the water so far as I am aware. But its presence has been clearly shown by an examination in this laboratory. "While working in the neighborhood of the lake recently it occurred to Mr. Gardner, of this division, to test the alkalinity of the lake water by adding a few drops of alcoholic phenolphthalein solution to a sample. No alkaline reaction could be observed, but on throwing away the tested sample and rinsing the containing vessel with distilled water a distinct pink color, denoting an alkaline reaction, was observed in the wash waters. An examination of the distilled water failed to show any alkalinity and the matter was referred to this laboratory. Samples of the water from the lake were examined and found to show no reaction with phenolphthalein, but on gradually diluting the solution the pink color appeared. Strong artifi- cial brines of sodium chlorid were then prepared and small quantities of sodium carbonate added. These also failed to show any color on the addition of phenolphthalein, but yielded the color on dilution. The reverse experiment was then tried of gradually adding sodium chlorid to a solution containing a small amount of sodium carbonate and col- ored a strong pink with phenolphthalein. As the concentration of the solution with respect to the sodium chlorid increased, the color gradu- ally disappeared. The explanation of the phenomena was clearly dem- onstrated. The small amount of carbonate present would be but little dissociated under any circumstances, the carbonates all having a very low ionization constant. But the very large number of sodium ions in the solution derived from great quantities of sodium chlorid' and sulphate present, the solution, in fact, practically saturated with respect to them, forced back the dissociation of the weakly dissociating sodium carbonate until there was none of this salt ionized in the solu- tion. Sodium carbonate itself is not capable of an alkaline reaction and its solutions only become so through its ionization products, as will be explained -later. In consequence this delicate test for its i^resence of necessity failed, and the small amount of carbonic acid involved could be very readily overlooked iu the analysis, especially when there 150 APPLICATION OF THEORY OP SOLUTION. was no reason to expect its presence. It may be of interest to note that an examination in this laboratory has shown the amount of sodium carbonate present to be about 0.012 per cent of the solution. The absence of more than a comparatively insignificant amount of lime from the waters of the lake has been the subject of much comment, especially as salt lakes in general contain a good deal of lime, some- times even an astonishing amount. That calcium salts have been brought into the lake in very considerable quantities there is most abundant evidence in the large amounts of carbonate of lime found on the shores and bottom of the lake. But calcium carbonate could not remain in solution in the lake water to any greater extent than a mere trace, for it is a very slightly soluble substance in itself and could not be ionized, as we have just seen, since the conditions are impossible for the existence of any carbonic-acid ions. Moreover, the water of the lake contains so much sulphate in solution that it must be regarded as practically saturated with respect to sulphions. Therefore only so much calcium sulphate can remain in solution as will be soluble without ioniza- tion — a comparatively small amount, as is well known. Should there be other calcium salts in the solution they would be continually dissociat- ing and forming calcium sulphate and being precipitated from the solu- tion until equilibrium should be restored. It should be remembered, however, that the soluble carbonates constantly being brought into the lake will tend to restore the sulphions to the solution, the carbonic-acid ions going out of solution with the lime. Nevertheless, it is the rela- tively large amount of sulphates of the alkalis present which must be regarded as the controlling factor in the small amount of lime soluble in the waters of the lake. TWO OR MORE SOLUTES WITH NO COMMON ION. Suppose, on the other hand, that to a solution containing an electro- lyte another containing no common ion be added. As has been already illustrated, there will be at once the formation of all the combina- tions possible to the ions, and with the consequent taking away of the ions for this purpose there will be a lowering of the product of ionic concentration for each of the salts; that is, the solubility will be increased. This phenomenon is most strikingly illustrated in the case of those salts usually regarded as insoluble. For instance, gypsum, which is essentially the salt calcium sulphate containing some water, is sparingly soluble in water. But the addition of an electrolyte with no common ion, such as sodium chlorid, will considerably increase the solubility of the gypsum. Some experiments made in this laboratory have shown that in moderately strong brines containing only sodium chlorid gypsum can be regarded as a soluble salt. The reason for this is readily seen when the substances which are formed are considered, both the calcium chlorid and the sodium sulphate being very soluble salts. The transportation of large quantities of lime by the drainage and HYDROLYSIS. 151 ground waters in arid regions where these salts are found is readily explicable from this point of view. Calcium carbonate, so abundant and so important in nature, is dis- solved in a precisely similar way; but the ionization of carbonates being relatively small, the effect is not so striking and relatively much less lime is transported in the solution. Tread well and Eeuter^ have recently published investigations on this point and find the solubility of calcium carbonate in sodium chlorid solutions does not become markedly large until considerable concentrations of the latter salt are reached. The effect of carbon dioxid in forming the more soluble bicar- bonate of lime undoubtedly is an important element in this connection, but as the ionization is but little affected by its presence its influence must be small in the presence of such a salt as sodium chlorid. HYDROLYSIS. Water is itself a very weak acid with a very low ionization constant, but it does dissociate to a limited extent into hydrogen and hydroxyl ions, thus: + - HaO t H + OH. The absolute amount of this dissociation is very small, but it has been the subject of the most careful and painstaking investigation. From the work of Kohlrausch, Ostwald, Arrhenius, and Weis it would appear that absolutely pure water would contain certainly less than 1 gram of hydrogen ions in 12,000,000 liters of water, but nevertheless there would always be some present. These dissociated ions from water will of course react with the ions of the dissolved electrolyte, thus: NaCl+HOH ^ NaOH+HOl, but the ionization constants of the sodium hydrate and the hydro- chloric acid are very large, while that of water is relatively very small indeed, so that the latter will be formed and precipitated as such nearly completely, and the total effect will be so small as to be incapable of detection by ordinary methods. This will be true for any electrolyte with high ionization constant, but with an electrolyte which ionizes to a small degree the effect becomes important. For instance, sodium carbonate has a very low ionization constant in water. The reaction is indicated thus : NajCOs+HOH t Na OH-f HNaCOj, and while the acid sodium carbonate does dissociate further it is to a very slight extent, as is the dissociation of the sodium carbonate and water as compared with the dissociation or ionization of the sodium hydrate; in consequence, the solution behaves as though it were, as in ■Zeit. Anorg. Chem., 17; 170 (1898). 152 APPLIOATION OF THEORY OP SOLUTION. fact it is, a solution of sodium hydroxid. This phenomenon of the dissociation of the salt of a weak acid or a weak base in water, with the formation of the corresponding strong base or acid possible, has been ca,lled hydrolhation. The ionization of water is increased by- heating, a fact of importance in many analytical operations. It can be very beautifully illustrated with calcium carbonate. If powdered marble be shaken up with water it dissolves to a very slight extent and is hydrolized with the formation of calcium hydrate, which can be shown by the alkaline reaction with a few drops of phenolphthalein. On heating the color becomes very greatly intensified, but on cooling it gradually returns to its former condition. This increased hydrolytic power of heated waters is obviously of the first importance in studying geological problems involving the solution of such weak salts as car- bonates, silicates, aluminates, borates, etc., but its full significance can hardly be said to be appreciated as yet. HILGARD ON THE R6Lla OF CARBON DIOXID. The imporfance of these relations in studying soil conditions and drainage waters, especially in arid regions where soluble salts abound, has not been .without some recognition. Hilgard, with Weber,' and later with Jaffa,^ has given the subject a good deal of attention and has formulated an hypothesis to account for facts observed. The impor- tance of the subject and the authority which attaches to the views of this investigator both call for a somewhat detailed examination of the hypothesis. From long observation Hilgard was impressed by the very great fre- quency of the occurrence of gypsum in alkaline waters and soil leach- ings, even when considerable quantities of soluble alkaline carbonates were present and when, a priori, the precipitation of all (practically) of the calcium in the gypsum as insoluble calcium carbonate might be expected. On looking over the literature he found a few sporadic references pertaining to the subject, wherein no explanation is advanced, or the phenomena are ascribed to some peculiar solvent action of the salts, without, however, any hypothesis as to the nature of this reaction. These solutions containing the gypsum or in contact with it are always alkaline, but on being evaporated to dryness or to relatively high concentration, this alkalinity is diminished and a greater or less amount of the calcium of the gypsum is precipitated as the ordinary carbonate. Considering the origin of these solutions, Hilgard is led to the view that the unlooked-for phenomena presented are due to the presence of an excess of carbonic acid in the waters or soil leachings, and points 'Proceedings of the Society for the Promotion of Agricultural Science, 1888, p. 40. = Ibi(l, 1890, p. 80. HILGARD ON THE r6lE OP CARBON DIOXID. 153 out the great probability of its presence. This view was tested by passing carbon dioxid, for various periods of time from ten minutes to two hours, into solutions of alkaline sulphates and chlorids, at the ordi- nary room temperature, in which precipitated calcium carbonate was suspended. Taking care to avoid the first masking of the reaction, which sometimes occurred from the large excess of carbon dioxid, in all cases the solutions were found to have become alkaline and, on the addition of alcohol to the solution, a precipitate was thrown down in which gypsum and calcium carbonate could both be readily recognized. Together with Weber he then undertook some quantitative measure- ments. He made up solutions containiug from 0.25 grams to 2 grams potassium sulphate per liter. In these solutions he suspended calcium carbonate and kept the solutions agitated while passing in carbon dioxid for 40 minutes. The solutions were kept at 18° during the experiment and a little litmus was added. At first the solution would redden, owing to the action of the carbonic acid on the litmus, but in about ten minutes it would become blue, the color intensifying as the reaction progressed. At the end, aliquot portions were filtered and tested as follows : I. The "total alkalinity" was determined by titrating with normal sulphuric acid (alkalinity=I^o. of cc's of the normal acid required). II. The portion was completely evaporated and dried at 110°, then leached, and the alkalinity of the leachings determined by titration with normal sulphuric acid. III. To this portion alcohol Vas added to make about a 60 per cent alcohol solution, whereupon a gelatinouspreci^itate settled, which, after twelve hours' standing, had separated into easily recognized crystals of gypsum and calcium carbonate. From the figures obtained it was concluded that to a point somewhere between the solutions containing one-half gram and 1 gram per liter of potassium sulphate the reaction was complete as to the replacement of the sulphion by the OOrion, but beyond this the replacement was less com- plete and the relative amount of the replacement was a regular func- tion of the initial concentration; further, that the solutions did not necessarily become saturated with respect to acid calcium carbonate, or the composition of the residue aifect the alkalinity. The further discussion by Hilgard, while interesting in itself, is not important for our purposes, and the examination of his views may fairly be undertaken with this statement of them. It should be remembered that calcium carl^pnate is itself soluble to some extent and hydrolized, as has been indicated on page 18, thus: CaOOs+HOH ^ 0a(OH)2+CaH3(CO3)2 and the solution behaves as a very weak solution of the calcium hydrate, itself but a slightly soluble substance, but distinctly alkaline toward indicators. 154 APPLICATION OF THEORY OP SOLUTION. When the potassium sulphate is brought into the solution, the SiOlu- bility of the calcium carbonate is materially increased, since the amount of the calcium that can be held in solution ascalcium sulphate is decidedly greater, and the equivalent amount of carbonic acid is now in the form of the very soluble salt, potassium carbonate, thus : KaSOi+OaCOs,:^* K2C03+CaS04 and the potassium carbonate formed being at'once hydrolized for the reasons already given, K2OO3+HOH ^ HKCO3+KOH, whereby we have the very soluble and highly ionized potassium hydrate and a characteristic strong alkaline reaction. It should be borne in mind in experimenting that even with the most iinely divided calcium carbonate some little time is required for the solution to come to equilibrium and this full effect to become observ- able. If carbon dioxid be now passed into the solution, that is to say, car- bonic acid be dissolved therein, at least two effects must be considered. It will tend to decrease the solubility of both the calcium carbonate and the potassium carbonate by driving back the dissociation, supposing, of course, that it is present in excess. But as the dissociation of these salts is very small, this effect of Its presence will not be appreciable, for its other result will be the formation of bicarbonate of both calcium and potassium, thereby increasing the solubility of the calcium and increasing the amount of this metal in the solution. The potassium bicarbonate is less soluble than the carbonate, but both salts are so soluble that this is no factor in the case before us. In spite of a widespread imiiression to the contrary, acid potassium carbonate has been clearly shown to be neutral and not alkaline in its reaction. This fact has been confirmed and most clearly demonstrated by recent work in this laboratory. But it has also been shown that it is a very instable compound, especially when in solution, quickly invert- ing with the formation of the normal alkaline carbonate, even though some carbon dioxide must be present. It is this inverted normal car- bonate which in turn is hydrolized with the formation of potassium hydrate, which gives the solution its alkalinity. The solution, after the treatment described, will then contain calcium sulphate, potassium sulphate, calcium carbonate or bicarbonate, potas- sium bicarbonate, and potassium carbonate hydrolized, and these constit- uents are in such proportions that stable equilibrium exists. As evap- oration proceeds this equilibrium is disturbed and the relative propor- tions vary, accompanied finally with successive separations of these various salts in the solid form. But always, up to complete desiccation, there will theoretically, at least, be some of each of these salts in the solution in equilibrium with the others, though the absolute amount of any of them may, and undoubtedly for some will, become quite small. GYPSUM AND AMMONIUM SALTS. 155 The addition of alcohol to the solution, by decreasing the solubility of some of the constituents, or all of them, will produce analogous results, differing only in the quantitative relations to one another. Thus the solution, it is obvious, must remain alkaline. After desiccation re- solution would restore the original conditions for the same total con- centration, if other things were equal. But it seems probable, from what we now know, that the solubility of the gypsum is materially affected by drying at 110°, and undoubtedly the acid potassium carbon- ate would be, which facts are quite competent to account for Hilgard's results with the leached residues. It will be shown later that the method he used for determining the alkalinity of his solutions gives misleading results, and his conclusions derived therefrom cannot fairly, be given more than a qualitative value. The main thesis of the paper, that the carbon dioxid dissolved in the soil and other natural waters is the controlling factor, at least in the sense he means it must be regarded as incorrect. The observed phe- nomena are readily accounted for in a much more complete way in terms of the conventional theory of solutions, and the presence of the carbon dioxid Is an unnecessary, though contributing, cause for the observed phenomena. GYPSUM AND AMMONIUM SALTS. It seems pertinent here to call attention to another reaction which has a widespread application. When ammonium carbonate and gyp- sum are brought into contact we may indicate the reaction that takes place thus: (NH4)20O3+ CaS04 1 Ca0O3+ (NH4)2S04, calcium carbonate and ammonium sulphate being formed. This fact is made of use for the preservation of ammonia salts in manure piles by the application of gypsum and, under the conditions that there exist, the reaction probably "runs to an end;" that is, all the calcium is con- verted into the carbonate. The same reaction results may be assumed for the bicarbonate of ammonium. Unlike the corresponding sodium and potassium salts, ammonium carbonate is quite instable. In water solution it immediately commences to break down, forming the acid or bicarbonate. This transformation is greatly accelerated by heat. The experiment was made in this laboratory of boiling a solution of ammonium carbonate for about twenty-five minutes, in contact with an excess of pure pulverized gypsum. The solution was then allowed to cool and settle. A major portion of the clear supernatant liquid was evapo- rated until crystallization. The crystalline mass was identified as am- monium sulphate, and on platinum almost completely volatilized when heated, a very small residue being left which could be recognized as a calcium compound by the flame test. As some of the ammonium sulphate must always remain in the solu- tion, even to complete dryness, it is obvious that the use of ammonium carbonate solution for the reconversion to carbonates of the alkaline 156 APPLICATION OF THEORY OF SOLUTION. earths after ignition, is not permissible when the soil contains gypsum or similar compounds. In spite of a warning by Hilgard' this fact is not always recognized and has been the occasion of some annoying experiences. AMELIORATION OF BLACK ALKALI. The amelioration of the conditions when "black alkali" exists is a subject for grave consideration. A suggestion has been made by Hil- gard which has been found to have great value. It is to apply gypsum or land plaster to the soil containing sodium carbonate. The reaction takes place as here indicated : NajOOj + CaS04 ^ OaCOj + Na^SO^, the noxious sodium carbonate being to a greater or less extent con- verted to less harmful sodium sulphate or "white alkali" and the harm- less and but slightly soluble carbonate of lime. The solubility of the calcium carbonate being so very small and made even less so by the presence of the more soluble calcium sulphate with a common ion, the tendency of the reaction will be to " run to the end " before equilibrium is established. But to bring about this desired result it is necessary that the soil should be well drained and aerated, partly, in order to carry oif the soluble salts ; and this may well be of considerable impor- tance, for there is evidence for the opinion that soils show a selective absorption or adsorption, and recent work of Mr. L. J. Briggs, of this division, indicates that sodium sulphate is much more readily leached from a soil than sodium carbonate. But more especially because in the presence of water and in a soil which is not well aerated there exists a condition favorable to the conserving of carbonic acid in the soil and, under the influence of this agent, the calcium carbonate is converted into the acid carbonate, its solubility much increased, and therefore its active mass, with the result that a reaction opposite to the one just indicated will prevail. Thus OaOOa+H^COs i^ CaH2(C03)2, and Na^SO4+CaH2(CO3)3?^0aSO4+NaHCO3, but the acid sodium carbonate formed, as will be shown, is instable in solution and at once inverts with the formation of the undesirable sodium carbonate, even when some carbonic acid must be present. NaHCOj:^ Na^COj-f H2O+OO2 Empirically these controlling conditions were determined, and pointed out by Hilgard. A correct understanding of their influence and impor- tance must prove of use in their management. ' Loc. cit., 1888, p. 40. ANALYTICAL PROBLEMS. 157 ANALYTICAL PROBLEMS. The preparation of a sample of soil or soil crust containing large amounts of these soluble salts presents peculiar difBculfcies. The prob- lem is not the complete chemical analysis of the specimens. This would but require the easy application of familiar conventional methods. What is desired is information as to the portion of the sample which is soluble in water. The total amount present in solution in water in contact with the soil can be determined in such cases where it may be desired by the investigator in the field, with altogether sufficient accuracy, by means of the well-known electrical instrument designed in this division. The question to be answered in the laboratory is what salts are present and their relative proportions in these soil solu- tions. The sample, as it comes to the analyst, is usually a dried portion of the soil or soil crust. Since the total amount of salts present is not desired, any convenient amount may be taken for the analysis. Two modes of procedure suggest themselves : I. To leach out the soluble constituents. Experience has shown this to be an ineffective method. It has been known since Bronner's work, in 1836, confirmed by the work in Liebig's laboratory, that of Way, and many others since, that soils are peculiarly adapted to the illustra- tion of the little understood phenomena of selective adsorption and absorption. Some little attention has been given to this subject in recent years, principally by Ostwald. But it must be said in all frank- ness that our knowledge of the matter is in a very unsatisfactory state. Attempts were made to extract some of these soil crusts by throwing them on a filter and extracting with successive portions of water. The mechanical difficulties in the filtering, the inefficiency of the method, as evidenced by the very considerable amounts of chlorid in the wash water after many extractions, caused this procedure to be abandoned. Some of the same soil crusts were then put in a continu- ous extraction apparatus which was known to be of a very efficient type. After extraction for three working days of eight hours, it was found that there were yet very large quantities of chlorid in the crust. After seven days' extraction chlorids were still coming through. Con- sidering the nature of the salts present, an application of the laws of solution seemed to cast doubt on the validity of this method, aside from the difficulties just described, so that it has been definitely aban- doned in this laboratory. It would seem desirable that a more system- atic and careful investigation of the matter should be made, and it is hoped that the subject will appeal to some one whose time and oppor- tunities will permit it. II. The procedure we follow at present is to take about 20 grams of the sample, if it appears to be largely soil, more or less according to judgment as the proportion of salts varies, and shake up thoroughly 158 APPLICATION OF THEORY OF SOLUTION. with about 35(1 cc. of water. This shaking or stirring is repeated fre- quently for a day. The whole is then allowed to stand overnight, or a day, as convenience dictates. The supernatant liquid, which is usually quite clear, is then decanted through a filter, as much as pos- sible being poured from the residual mud without actually draining it. The filtrate is then made up to about 500 cc. and aliquot portions taken for the analysis according to the conventional method. Certain assumptions are made in this procedure, but they appear warrantable. It is to be supposed that some of all the salts remain in the residual mud. Gypsum, which is a very common constituent in these samples, undoubtedly fails to be extracted to anything like completeness. More- over, the solution which is decanted from the soil may well be quite different in concentration from the solutions which would occur ia the field, and consequently the proportion of salts dissolved may be differ- ent from the larger volume of water used. Taking these facts into consideration, it nevertheless seems probable that the solution as we j)repare it gives a fair approximation to the soluble constituents of the soil, barring, of course, excessive amounts of such a substance as gyp- sum, and will give a fair idea of what a drainage water in that soil would carry away. We can only say that in the light of our expe- rience, and after giving a great deal of consideration to the subject, it seems to present fewer objections than any procedure so far suggested. It sometimes happens that after decanting through the filter the solution will remain more or less cloudy, even after prolonged stand- ing. When cloudiness is due to suspended mineral matter, it can gen- erally be removed by running the solution a second time through a Schleicher and Schiill "hardened" filter, drying the filter, and washing as quickly as possible with several successive small portions of hot water. In general what the analyst has to estimate in the solution are the chlorids, sulphates, and carbonates of calcium, magnesium, and sodium. Occasionally other substances, as the, salts of potassium and strontium, may be present in sufficient quantity to require their estimation, but this is exceptional; usually they are present in such minute quantities that they may be disregarded. Too little is known, as yet, regarding the nature of the organic matter which may be in the solution, and its effect on the other constituents to make its determination of any value. It is important, however, that a direct determination of all the constit- uents usually found should be made, and it can not be assumed that any one of them, the sodium for instance, can be calculated by differ- ence. If there be organic matter in the solution, an attempt to deter- mine the total mineral constituents will necessitate its biirniug at a temperature which will inevitably cause a loss in the other constituents. Its oxidation by other methods, while possible, is not justified on account of the time required, and, moreover, either liberates the car- bonic acid or i^resents other chemical objections. The diflBculty in ANALYTICAL PROBLEMS. 159 exactly dehydrating a residue, when there will be so much "water of crystallization," will be sufficiently obvious to any analyst. To attempt to calculate the amount of sodium required with the other bases to exactly neutralize the acid found will often yield misleading results where there are mixtures of monovalent and bivalent acids. Under such circumstances different analysts would get different results, depending upon the way in which the various bases and acids were combined. The statement of the analytical results merits attention. The con- ventional method of stating them as salts is often misleading and gen- erally quite unwarrantable, with our present knowledge. To illustrate, the writer had occasion recently to compare the analyses of the mineral constituents of the Great Salt Lake, made through a wide range of time. The analyses were all calculated in the same way, presumably for pur- poses of comparison, and are stated thus : Gale, 1850. Allen, 1860. Bassett, 1873. Taliiiage. Sodinm cMorid Sodium sulphate Magnesium chlorid . Calcium sulphate - . . Potassium sulphate . Potassium chlorid . - Excess of chlorine . . 20.20 1.83 .25 Total. 11.80 .93 1.49 .09 .53 8.85 1.09 1.19 .20 .09 l.S 13. 686 1.421 1.129 .148 .432 13.42 16. 716 15. 743 1.060 2.011 .279 .474 19, 557 l^Tow the chlorine uncombined, stated in several of these analyses, can all be accounted for by recombining the data in a different way than that stated. It is not necessary to go into the details, as it is obviously a matter of mere arithmetical jugglery. It will be observed that all the magnesium is stated as having been present as chlorid, whereas it is probable that many other analysts, perhaps the majority, would have calculated it as sulphate. As to which would have been the better way there can be no dispute. By referring to the discussion in the earlier part of this paper it will be seen that as a matter of fact there must necessarily have been both magnesium chlorid and magnesium sulphate in the solution. In just what proportions they would be it is of course possible to say theoretically, but the practical difficulties in deternjining this question, in the vast majority of cases, may be regarded as at present entirely too difficult to warrant any attempt to do so. In view of the fact that it is the ions with which we really have to deal in the majority of instances, and that it is- in fact the amount of the possible ions which may be formed that we actually estimate in our analyses, it would seem that the time has come when some institution or organization that could speak with authority should inaugurate the convention of stating analyses in this more rational way. The diffi- culties besetting any individual or sporadic attempts in this direction 160 APPLICATION OV THEORY OF SOLUTION. are so patent they need no discussion. Yet the need of this reform for an intelligent appreciation and use of the material can not be too strongly urged. Since the epoch-making work of Liebig there has been a constant effort on the part of chemists to determine the agricultural value of a soil by a chemical examination or analysis. The methods which have been suggested all depend upon the principle of bringing the constitu- ents of the soil into solution to a greater or less extent in the presence of some acid. The constituents of the soil are for the most part salts. These salts are in general but slightly soluble, it is true, but nevertheless soluble to some extent; and consequently in aqueous solutions would be ionized as other salts are, or, since they are usually salts of weak acids, hydro- lized, as has been described elsewhere in this paper. The presence of an acid, or, for that matter, any other electrolyte, such as a base or a salt, displaces the equilibrium and alters the solubility of the soil con- stituents. Acids have met with most favor in this regard, and the reason for this is now apparent, because in dissociating they do not furnish ions common to the dissociating soil constituents, and hence the solubility changes are more marked when they are employed. It will be readily seen that these acid extracts represent a state of equilibrium between the dissociated and undissociated electrolytes contained in the solution. If the acid solution be allowed to remain long enough in contact with the soil, it will become saturated with respect to the soil constituents or their ions if they are present in sufficient quantity, or will completely dissolve such of these constitu- ents as are not present in sufflcient quantity to saturate the solvent with respect to them; but with salts such as compose the soil the surface exposed as controlling the " active mass," as well as their com- paratively small solubility, must be considered, and for a final com- plete state of equilibrium to be obtained will require generally that the solvent be kept in contact with the soil for a relatively long time. It is obvious that in any particular case many factors or conditions enter, and the subject is bewilderingly complex. The temperature as controlling the solubility of the various electrolytes themselves; as controlling the migration velocities of the ions; the concentration of these various electrolytes and their resulting ions; their mutual effect in retarding or increasing solution when they yield a common ion or do not yield a common ion; the surface exposed and the time which it is exposed, must all be considered. It is not surprising, therefore, that experience has shown different acids to have yielded widely varying results; that the same acid would not always yield comparable results, even when the same con- centration was used, for in such a case the factors, common ions, and active mass are incontrollable. The practice finding most favor at the present day is to extract the soil with a hot concentrated aqueous sola- FIELD METHOD. 161 tion of hydrochloric acid, but to this procedure there is the objection that the concentration with respect to hydrochloric acid, and therefore the active mass of this reagent, is measurably dependent upon the barometric pressure. The conditions which will determine the equilibria in our acid extracts are similar in kind to those which control the equilibria of our soil solutions, but they are so utterly incomparable in degree as to make any attempt at formulating a quantitative comparison well-nigh hopeless. And when we consider the effect of other phenomena which are known to come into play in the soil, such as absorption or selective absorption, for instance, this attempt must be regarded as absolutely hopeless until we shall have obtained much more and precise knowledge regarding these subjects. CARBONATES. It has often been found desirable in our field work to obtain a knowl- edge of the relative quantities of the various salts which may be present. Reference of samples to the laboratory for analysis by the regular methods is not practicable, on account of the time involved, it being necessary that the desired information be obtainable within a few hours at the furthest, in order to determine the direction in which the work in hand should proceed. To meet these requirements the following scheme has been devised : The apparatus has been reduced to a minimum ; it has been so arranged that it can be readily carried into the field, and our parties are regu- larly provided with a small carrying case, in which everything needful in the analysis is securely and neatly packed. It is provided with four or five burettes of 25 cc. capacity, graduated to iV cc. (^ cc. will answer every purpose quite as well) j several bottles (generally four) holding 250 cc. when filled to mark on the neck ; specimen vials and small bottles, to carry reagents — the solid reagents required for the standard solu- tions are carefully weighed into specimen vials or tubes in the labora- tory, and sent out as required, the solutions being made in the field by means of the 250 cc. bottles mentioned ; a small porcelain casserole, in which the titrations are made, and which can be heated over any fire — sometimes small Erlenmeyer flasks and a spirit lamp are preferred; and several clamps for fastening the burettes to the side of a wagon. It is obviously impossible to provide a complete analysis under the conditions prescribed, but for the purpose desired it is sufficient to know the relative amounts of carbonates, sulphates, andchlorids present. Therefore the method is designed to estimate these acids, which for convenience are conventionally expressed in terms of the sodium salt. A sample of water is taken, or an extract is made from the soil, and the solution filtered or decanted. The solution need not be clear. H. Doc. 399 11 162 APPLICATION OF THEORY OF SOLUTION. A known amount of barium nitrate is added in excess to 10 cc. of the salt extract, to precipitate sulphates and carbonates. The excess of barium nitrate is titrated back with a solution of potassium chromate, using silver nitrate on a porcelain tile or plate as an indicator. In the same vessel standard silver nitrate is run in to precipitate the chlorid, using potassium chromate on the porcelain plate as an indicator. A few drops of nitric acid are added, and the liquid heated to boiling, to drive off the carbon dioxid precipitated with the barium. The excess of nitric acid is neutralized with powdered magnesium carbonate. Again the solution is titrated with potassium chromate, the quantity required indicating the amount of carbonate in the original solution. This, sub- tracted from the amount required for sulphates and carbonates, indi- cates the sulphates present in the solution. The three titrations are all made in the one vessel, making the method a very rapid and compara- tively simple one. The method is not without its shortcomings, but it is in the nature of things intended only to obtain approximate results and has proven itself well adapted to this purpose. It may well be capable of refinement. It should be noted that the estimation of the barium by adding potassium chromate, using the silver nitrate as an indicator, can not be advantageously done by adding the silver nitrate to the solution, and even on the porcelain plate the reaction is not all that could be desired. The presence of a chlorid in the drop taken for the test necessarily masks the reaction undar observation, and even in the absence of chlorids the reaction is not "sharp." The writer is indebted to Prof. J. D. Tinsley, of New Mexico, for assist- ance in testing this method, with a view to modifying it for laboratory practice. Without going into details, it may be said at once that the reaction does not appear well adapted to such an application of it. Its use in the field method just described is, however, warranted, con- sidering the nature of the results desired. The scheme, as described, has been found of service for the purposes for which it was designed. ESTIMATION OP CARBONATES AND BICABBONATES : A FIELD METHOD. In the course of their survey work in the regions where the "black alkali" exists, our field parties felt the necessity of a rapid method for the estimation of the soluble carbonates in the soil, or really the alkalinity of the soil. They tried at first to meet this want by taking a standard volume of the soil, shaking it up with water and titrating with a stand- ard acid. The results were very soon found to be valueless and the problem was referred to this laboratory as urgent. Obviously, the first thing to do was to study the reaction between an acid and sodium carbonate. It was very soon found that while very good concordant results could be obtained by heating to boiling the solution of sodium carbonate before titrating with standard sulphuric acid solution, that FIELD METHOD. 163 such was far from the case when working in the cold. It ultimately developed that the formation of the acid carbonate or bicarbonate of soda from the liberated carbonic acid was the cause of the trouble. This salt, in spite of a widespread impression to the contrary, is per- fectly neutral toward the usual indicators, but is very instable, and inverts readily to a greater or less extent, depending on the conditions, to the normal carbonate which is alkaline in water solutions. In titra- ting, therefore, with a standard acid solution widely varying results have been obtained, depending upon the rate at which the solution was run in, the temperature, etc. In order to test this reaction more thoroughly a solution of sodium carbonate was prepared. Fifty cc. of this solution, treated with an excess of barium chlorid solution, gave 0.4840 grams of barium carbonate, equivalent to 0.2601 grams of sodium carbonate. Therefore, 1 cc. of the solution contained 0.005202 grams of sodium carbonate, while a tenth normal (N/10) solution should contain 0.00528 grams of the salt. A solution of sulphuric acid was made up and, after a number of pre- liminary trials to obtain concordant results — without success — a titra- tion was made in the usual manner, using phenolphthalein as the indi- cator and titrating to loss of color, the solution being run in from the burette fairly rapidly. 10 cc. of the sodium carbonate solution was equivalent to 1.38 cc. of the acid solution. Therefore, 1 cc. of the H2SO4 solution was equivalent to 7.2 cc. of the KazOOs solu- tion, and 7.2 cc. of the Na2C03 solution was equal to 0.0374 grams of sodium carbonate. If the reaction is to be represented thus : Na2C03+ H2S04=Na2S04-f H2OO3, then 0.0374 grams of sodium carbonate should be equivalent to 0.03456 grams of sulphuric acid. If the reaction is to be represented thus : 2 Na2C03+H2S04=2 NaHCOj-f ]Sra2S04, then 0.0374 grams of sodium carbonate should be equivalent to 0.01728 grams of stilphuric acid. But 50 cc. of the sulphuric acid solution yielded 2.3345 grams of barium sulphate, therefore 1 cc. representing 0.0374 grams sodium carbonate was equivalent to 0.0196 grams of sulphuric acid. From these results it was concluded that both reactions took place in this case, the latter predominating. Similar results, but different in the quantitative relations, confirmed this general conclusion. The method was therefore abandoned. Various methods were then tried and the 164 APPLICATION OF THEORY OF SOLUTION. literature tlioroughly scanned witliont finding anything suited to our purpose. Ultimately tbe following reaction suggested itself: XaiC03+HKS04=NaHC03+XaKSO,. Acid potassium sulphate is a well-characterized acid, while both the reaction products are neutral. A tenth normal (N/10) solution of the acid potassium sulphate was prepared and titrated against the sodium carbonate solution, already described, using phenolphthalein as the indicator. 10 cc. of the sodium carbonate solution required 9.9 cc. of the acid potassium in solution, and this result was obtained time and again, only on standing the solutions gradually regained their pink color, due to the inversion of the acid sodium carbonate with the for- mation of the sodium carbonate. While this inversion commences almost at once, it is not rapid enough to interfere with the titration. This point will be referred to again later in this discussion. It was proved that the reaction takes place as indicated with quanti- tative exactness. The reagent employed is one which can be readily obtained in any desired purity, and is very easy to handle. It remained to test it for field conditions. A solution of sodium carbonate was pre- pared, 10 cc. of which was equivalent to 13.5 cc. of the acid potassium sulphate solution. 50 cc. of this sodium carbonate solution and 10 grams of a loamy soil were put in a 500 cc. measuring flask, water added to the mark, and the whole vigorously shaken for several minutes. The solution was then allowed to stand and the soil to partially settle for five minutes. 100 cc. of the solution was then drawn out with a pipette, phenolphthalein solution added, and the turbid solution quickly titrated with the acid potassium sulphate solution, requiring 13.5 cc. This experiment was repeated with identical results. In order to test the effect of sodium bicarbonate, if present in the soil, as we had reason to suspect was sometimes the case, a solution of this salt was prepared. It partially inverted, and blanks made with it showed that 10 cc. of the solution was equivalent to 0.8 cc. of acid potassium sulphate solution. Ten cubic centimeters of this solution were added to each of two 100 cc. portions of the muddy solution containing sodium carbonate, just described, and they were both titrated at once. They required 14.1 cc. and 14.2 cc. of the acid potassium sulphate solu- tion, respectively, or, subtracting the correction, 0.8 cc. for the bicar- bonate, 13.3 cc. and 13.4 cc. Therefore, it may be assumed that the presence of an initial amount of bicarbonate does not affect the reaction materially. In order to test the method for field use a strong alkali soil, known to contain considerable sodium carbonate, was exactly saturated with water — a process capable of great precision in the amount of water used. Two portions of 20 grams each of the saturated soil, which we will designate as A and A', were washed into 500 cc. measuring flasks, water added to mark, and the whole vigorously shaken. They were FIELD METHOD. 165 allowed to stand jast five minutes to settle somewhat, and then 100 cc. portions were drawn off with a pipette for titration. The whole process was then repeated with fresh samples of this same soil, which we will designate as B an dC. The nnmberof cubic centimeters of a -^ HKSO4 solution required are here given : A. A'. B. C. 49.1 48.8 48.6 48.7 50.3 50.0 49.2 49.2 49.1 48.0 49.0 48.8 45.8 50.9 46.3 48.7 Average, 48. 8 49.7 48.7 47.9 Under the conditions of the experiment these results may be regarded as quite satisfactory. They were determined by two observers indiscrimi- nately and as a first test of the method. The particular soil chosen was one which gave a dark gray mud whose color made the end of the titration unusually dif&cult to determine, and, finally, all these determi- nations were made in a very bad light. Efforts were next directed to perfecting the mechanical details for field use. It may be said at once that the method has received a most rigid testing in the field as well as the laboratory, and on solutions of known strength as well as on unknown soils, and has been found to be of easy application and very accurate. In the absence of any very unusual special condition there is no difficulty to titrating accurately to within 0.1 cc. of the K'/IO acid potassium sulphate solution. This would correspond to 0.000528 grams of sodium carbonate. Supposing we started with 20 grams of satu- rated soil containing 15 grams of solid constitutents and made up the solution to 500 cc, titrating with 100 cc. portions, this would mean an error of 0.017 of 1 per cent. In soils containing from 0.5 per cent to 4 per cent of sodium carbonate this is an accuracy far beyond field requirements. In this laboratory the method is always used in the examination of alkaline soils and waters, and of course with the greatest accuracy that the circumstances will justify. It possesses the inesti- mable value of showing at once just how much of the sodium carbonate is present as the normal salt, capable of being hydrolized, and acting as an alkali. For this purpose, the usual method of titrating with a standard sulphuric acid solution or of estimating the sodium or carbon dioxid can only be regarded as entirely inadequate and misleading. The apparatus and procedure finally adopted for the use of our field parties is as follows : Several (usually four) glass bottles of convenient shape, with cork stoppers, of such a size as to hold 250 cc. of water at 20° C. when filled to a mark on the neck ; one or more 25 cc. burettes, graduated to iV cc. (i cc. would amply fill all requirements) ; a stand or clamp to attach burette to side of wagon or box; one or more 50 cc. pipettes; a vial of 166 APPLICATION OF THEORY OP SOLUTION. a concentrated alcoholic solution of plienolpbtbalein ; :i small vessel for measuring the soil. This may vary somewhat according to predi- lections of the user. Before going into the field the weight of a satu- rated soil, the corresponding weight of this same soil when dried at 110°, for soils of different textures, which this standard vessel will hold when exactly filled, is determined and the results are tabulated in the ob- server's note book. The texture of the soil, and in consequence the weights just referred to, can be determined by observation with remark- able accuracy after some jiractice in the field. Small glass vials, each containing 1.C94 grams of pulverized acid potassium sulphate, carefully weighed and prepared in the laboratory, are carried or sent to the field parties as they may .need them. This amount of the acid salt when dissolved in 250 cc. of water gives a F/10 solution, which is usually enough for a large number of determinations. The actual determina- tion is made by taking from the cell of the electrical instrument used for determining the total salt content of the soil enough of the saturated soil to just fill the standard vessel. This is then carefully washed into one of the 250 cc. bottles, in which process a small copper funnel is found very useful. The bottle is then filled to mark, shaken vigorously for a few minutes, and allowed to settle for five minutes. Fifty cubic centi- meters of the supernatant liquid is drawn off' by a pipette and run into a tumbler, Erlenmeyer flask, or any convenient vessel. A few drops of the plenolphthalein solution are added, and if it indicates the presence of an alkali the solution is quickly titrated with the acid potassium sul- phate solution. Supposing the apparatus and solution to have been assembled before starting on the day's work, this whole operation can be readily performed in from ten to twenty minutes, including the noting of the data. As five minutes of tliis time is available for other purposes, the method is undoubtedly a rapid one, and has proved from actual experience to be far more accurate than our requirements. In developing this method some observations were made which seem worthy of noting. Calcium carbonate is somewhat soluble in water and is hydrolized so that the solution becomes slightly alkaline. The calcium carbonate is converted into a neutral salt by the acid potas- sium sulphate, so that a false reading may be obtained in estimating the sodium carbonate. In the case of soils or soil crusts this is of no impor- tance, for it takes the calcium carbonate quite a little while to go into solution to a sufficient extent to make this effect noticeable, unless there be another salt such as sodium chlorid or sulphate present, when, as has been pointed out, one is entirely justified in regarding the alka- linity as due to sodium carbonate formed. In testing river or canal waters, however, which show a faint alkalinity, it is well to investigate whether the alkalinity may not be due to calcium carbonate alone. The question can usually be settled at once by testing for the presence of a chlorid or sulphate in the water, in which case when either of these is found the alkalinity may be safely regarded as due to sodium carbonate. INVERSION OP SODIUM BICARBONATE AND SODIUM BISILICATE. 167 This method can not be considered as available for the reverse pro- cedure — that is, to run the sodium carbonate from the burette into a solution of the acid potassium sulphate. This would be an advantage for some observers, if it were allowable, because one could then titrate to the appearance of color instead of to the disappearance of it. But acid potassium sulphate must be regarded as a strong acid, quite capable of decomposing carbonates, and by having it in excess, as has just been supposed, the acid sodium carbonate which would be first formed would then be more or less decomposed by the excess of acid, depending upon the temperature and concentration conditions, with the result that the readings obtained would inevitably be nonconcordant and a quantita- tive interpretation of them would be impossible. INVERSION OF SODIUM BICARBONATE AND SODIUM BISILICATE. The fact that acid sodium carbonate, or the bicarbonate, as it is often called, is an instable compound has been alluded to quite frequently in this paper. So, when solutions which contained the carbonate have been titrated to absence of color — that is, contain only the acid carbon- ate — are allowed to stand they gradually regain color, indicating the inversion of the bicarbonate and reformation of the carbonate. This has suggested a way of studying this inversion quantitatively, a knowl- edge of which would probably have great practical value as well as theoretical interest. This subject is now under investigation, but only preliminary results have been obtained as yet. It would appear that at the ordinary room temperature (18° 0., about) the inversion pro- ceeds quite rapidly, the curve being quite steep until equilibrium is nearly reached, when it suddenly becomes very flat and assumes an assymptotic character. This reaction has now been very thoroughly tested for both sodium and potassium carbonates, and in the presence of their respective bicarbonates has proved itself most satisfactory. Among the applica- tions of it which have suggested themselves is its use as a method for estimating the relative amounts of the carbonate and bicarbonate in the manufacture of soda and potash, for which up to the present time no satisfactory method has been found. It remains only to say a word as to the indicators. Other indicators have been successfully used, but phenolphthalein has proved by far the most satisfactory for the " sharpness" of the end reaction, the intensity of the color, and because one is not bothered by any change of color with mixed results, the color being entirely absent the instant the sodium carbonate ceases to exist as such. It has been suggested that carbonates and bicarbonates can be esti- mated in the same solution by using phenolphthalein as the indicator for the normal carbonates and as soon as color disappears, methyl orange, which is a stronger acid than carbonic acid, being added and the titration continued. Serious objections have been made to this 168 APPLICATION OF THEOKY OF SOLUTION. inetliod when the ordinary acids are used.' On the other hand it has apparently worked satisfactorily for the phosphates.^ Mr. Briggs has found that when potassium hydrogen sulphate is the acid employed, the method is quite satisfactory for the alkaline carbonates and silicates. It is only necessary to add a few drops of phenolphthalein titrate to loss of color, add a little of the methyl orange at once and continue the titration; the difference between the two titrations giving tlie equivalent of sodium bicarbonate originally in the solution. A more critical examination of the method is now in progress. It will be of interest to note in this connection some striking resem- blances in the conduct of sodium silicate to that of sodium carbonate. As the salt of a weak acid it is hydrolized in water solution and shows a strong alkaline reaction with phenolphthalein. Carefully titrated with acid potassium sulphate until color Just disappears, if allowed to stand it quickly shows a return of the color, which gradually becomes more intense. The explanation for this phenomenon is undoubtedly the same as for sodium carbonate. Further work along this line is in progress in this laboratory, and will be prosecuted as rapidly as cir- cumstances will admit. GENESIS OF HAEDPAN. The application of the present views regarding solutions to the study of hardpan phenomena gives promise of valuable as well as interesting results. A hardpan may be defined as a layer of the soil, usually near the surface, having the texture of the soil just above and below it, but more or less closely cemented by some material. In general, hardpan is a characteristic of soils where drainage is very poor or where stand- ing soil waters may accumulate. The cementing material is often lime carbonate, but may be other material, as the hydrates of iron and alumina or silica. Hardpans vary very much in their physical proper- ties. They are sometimes dense and close-grained as a well-character- ized rock, requiring blasting or similar methods to break them up. In other cases they may be partly porous, and when brought to the sur- face disintegrated with ease, and there are all grades between these extremes. The objections to their presence in the soil are evident. They pre- vent the penetration of plant roots, and, more important, they prevent the moisture from rain, irrigation, etc., sinking into the soil and thus being conserved for future use. They also prevent the water that may be beneath them from being drawn to the surface and made available for the plants. The formation of a calcium carbonate hardpan is the most readily understood, and this has been dwelt upon at some length in a paper by Gardner and Stewart. It is there pointed out that resolution and re- ' Kuster. Zeit. anorg. Chem. 13, 127 (1896), 2 Cavalier. Comptes rendus, 126, 1142, 1214, 1285. 127, 60 (1898). THE GENESIS OF HAKDPAN. 169 precipitation are important factors. But when the calcium carbonate does not exist, as such, in the soil or in the vicinity, so as to be brought by water, while a limestone hardpan might form, under favorable con- ditions, it seems more probable that the cementing material would be one of the other substances mentioned, or a mixture of them. The mineral constituents of the soil are for the most part salts, but with a few exceptions salts with a very limited solubility. Neverthe- less, to some extent at least they are soluble, as are other salts, and their solubility may be increased or diminished by the presence of another salt solute, as has been indicated in a former part of this paper. These salts, carbonates, silicates, aluminates, ferrates, etc., are without exception salts of weak acids and may be expected to be much hydro- lized in as far as they are soluble at all. This has been very beautifully illustrated in recent experiments by Clarke,' who has treated a large number of minerals carefully pulverized with pure water. On the addi- tion of a few drops of dilute alcoholic phenolphthalein a marked alka- line reaction could be observed in the great majority of the cases investi- gated. The reaction may be indicated thus, assuming a very simple example to exist : ESiOa+HOH ^ EOH+HzSiOs All these other substances are very slightly ionized in comparison with EOH. If R be a well-marked base, such as sodium or calcium, the solution will therefore be alkaline, as has been shown to be the case with calcium carbonates, sodium silicates, etc. The fact that the sili- cate is complex will not alter this general property. Precisely similar conduct is to be expected of aluminates and ferrates. This means that there will actually exist in the solution some of the hydrates of alumina, silica, or iron, as the case may be, which will remain as such on evapo- ration, though the absolute amount may be very small. The bases will be more or less readily removed, as they will be brought in contact with the carbonic acid and other acids (organic 1!) of the soil to form compara- tively readily soluble salts. This process probably plays an important part in the formation of bog-iron ore, which may be regarded as strictly analogous to a hardpan. The deposition of bauxite, for example, or the formation of a siliceous conglomerate is essentially of the same nature. But it should be remembered that in these latter cases when the action has been deep- seated with hot water as the solvent, the reagent has been much more ionized and so is much more efficient as a solvent. Au interesting case from southern California has recently come to our attention. The soil was shown to have been somewhat compacted under the plow sole. When the irrigating water was applied, this packed region of the soil caused a more or less temporary accumulation of the 1 Jonr. Am. Chem. Soc. 20; 739 (1898). 170 APPLICATION OP THEOEY OF SOLUTION. waters. This soil, as can be readily seen under the microscope, con- tains a large proportion of unaltered mineral fragments, rich in iron and alumina and therefore well adapted to yielding these materials under the influence of the solvent action of the water; and, as a matter of fact, this packed material is found to rapidly become cemented with iron and alumina, as an examination in this laboratory showed. It is to be regretted that at the time this examination was in progress it was not deemed expedient to determine what constituents the irrigating water held which might augment its solvent power. Thiit other agencies are at work in the production of these phenom- ena may well be the case. For instance, oxidations undoubtedly have a significant r61e in this connection in breaking up the original miner- als. But it seems equally certain that the part that solutions play has not been given the consideration that it merits, mainly because solution phenomena have not been understood until comparatively recent years. The study of hardpan formation necessitates a consideration of cer- tain physical phenomena; for instance, the movement of water and various solutions in the soil. This subject is receiving attention in this laboratory; but while a good many observations have been made and much valuable data collected, it is yet too soon to formulate a complete hypothesis for this subject. The views here described are put forward in the hope of furnishing an incentive to more widespread interest and work on this important subject. THE e6lE op fertilizers. In the action of fertilizers on the soils there is very much to be learned ; for, in spite of the great amount of attention that has been given to this important subject, it can not fairly be claimed that we have made much more than a beginning. While it has long been acknowledged that the disintegration of the mineral constituents of the soil, through the agency of the fertilizer, is a factor in its use, and that solution phenomena are imxiortant in this connection, but very little advance has been made in this direction, and its full importance has not been appreciated. It can not fail to be of advantage to inves- tigators in this field to give this point of view more consideration in the future; for it is indisputable that the mineral constituents of the soil are salts soluble to at least some extent, and affected in their solu- bility, as are other salts, by the presence and nature of other solutes in the soil solutions. It becomes of the first importance, therefore, in determining the use of any particular fertilizer to consider its probable conduct on contact with water — what ions it will yield, if it be an electrolyte, and how these ions will affect the solubility of the soil ingredients, and thus bring into activity products which may be of value as plant nutrients. No less desirable is information of this char- acter than is that of the physical effect as modifying the texture or moisture content, for exami)le, or in forming new compounds more or SUMMARY. 171 less desirable; and in the study of this latter point the ideas here sug- gested will be found to have much use. That they will lead to further work on the purely physical side of the problem is not the least of the advantages in their consideration; for it does not seem too much to hope that the near future will bring us most helpful and practical infor- mation as to selective absorption of solutes by soils and as to the change of vapor pressure in soil solutions, important as modifying the moisture content of the air in contact with the soil. Other questions of a kindred nature will suggest themselves, and the time is ripe for a larger, broader and more scientific attack on these important problems of agriculture. SUMMARY. In this paper there have been presented: 1. An outline of the theory of solutions, showing that a solute by virtue of the presence of the solvent behaves as though it were a gas, and that electrolytes present the added phenomena of electrolytic dissociation or ionization. 2. A demonstration that the reactions under investigation are of a reversible type, and in consequence the Mass law is applicable to a study of the equilibria among the solutes. 3. An application of these views, showing how the solubilities of the sulphate and. carbonate of lime in nature are increased by the presence of a solute which dissociates but yields no common ion. 4. An announcement of the presence of sodium carbonate in the waters of the Great Salt Lake, Utah, and an explanation of why this fact has previously escaped observation, based on the relation which obtains between the ionization products and the solution constant. A similar explanation is offered for the scant amount of lime in the waters of this lake. 5. An examination of the hypothesis of Hilgard as to the role of carbon dioxid in the genesis of alkali, in which it is demonstrated that the phenomena observed are more satisfactorily accounted for in terms of the theory of solution, and that the carbon dioxid must be regarded as a contributing cause, but not a necessary one. 6. An examination of the Hilgard method for the reclamation of black alkali soils, with an explanation of the reactions observed and of the importance of the controlling conditions respecting drainage and the accumulation of carbon dioxid, empirically announced by Hilgard. 7. An examination of the reaction between calcium sulphate and the carbonates of ammonium. The use of gypsum for conserving ammonia in manure piles is explained. Some errors', with the reasons therefor, which may accompany the use of ammonium carbonate in analytical operations involving salts of the alkaline earths are pointed out. 8. A discussion of some analytical problems in a chemical examina- tion of alkali soils. The nature of the problems is made clear. The 172 APPLICATION OF THEORY OP SOLUTION. relative merits of leaching the soils and taking a solution in contact with the soil in preparing the sample are discussed, and the advan- tages in favor of the latter procedure indicated. The necessity of mak- ing a direct estimation of each constituent is demonstrated. 9. A plea for the rational statement of analytical data, inasmuch as it is the ions which are determined and not the salts. Furthermore, it is the ions with which we are generally concerned in the study of any particular problem. 10. A iield method for the estimation of sulphates, chlorids, and car- bonates, involving three titrations which may all be made on the same sample in one vessel. Its use in reconnoissauce work is described. 11. A rapid method for the estimation of sodium carbonate in the presence of the bicarbonate, depending on the conversion of the alka- line carbonate to the neutral acid carbonate, with the formation of a neutral sulphate by the addition of acid potassium sulphate. The use of the method in the laboratory and in the iield is described, and its prob- able availability for technical work is suggested. The objections to the use of sulphuric acid in determining "alkalinity" are made evident. 12. Observations on the hydrolysis of sodium carbonate and sodium silicate and the inversion of sodium bicarbonate and sodium bisilicate to the normal salts are described briefly. 13. A discussion on the formation of hardpan and similar deposits, in which it is pointed out that the hydrolysis of the salts of weak mineral acids and subsequent desiccation and deposition of the solution products must be taken into account in any hypothesis as to their genesis. 14. An explanation of the solution and hydrolysis of certain minerals and the consequent alkalinity they display. 15. Suggestions for the study of the functions of fertilizers, in which the importance of considering the solution phenomena which their presence may effect in the ground waters is made evident. 16. Some observations on selective absorption and other physico- chemical phenomena which are incidental to a complete study of the properties of a soil. SOME NECESSARY MODIFICATIONS IN METHODS OF MECHANICAL ANALYSIS AS APPLIED TO ALKALI SOILS. By LYMAN J. BRIGGS. INTKODUCTION. Many alkali soils present peculiarities in composition which do not permit the use of the ordinary methods of determining their mechanical composition. It consequently seems desirable to call attention to cer- tain modifications in the methods of examination which have been found useful in determining the mechanical comiiosition of samples of alkali soils collected by the field parties of this division. The following points will be considered : (1) The disintegration of the soil during the progress of the analysis, resulting from the solvent action of the water used in making the mechanical separation. (2) Apparatus and method for examining soils subject to excessive disintegration during mechanical analysis, and the advantages of the centrifugal method as applied to all soils. (3) The treatment of the mechanical separations after ignition to convert the oxides of the alkaline earths into carbonates. (4) The determination of the water-soluble salt content of soils in connection with their mechanical analyses. MECHANICAL ANALYSIS OF SOILS SUBJECT TO EXCESSIVE DISINTEGRATION. Many of the soils of regions requiring irrigation contain consider- able amounts of gypsum (CaS04+2H20) and calcium carbonate. In many cases these substances will be found cementing together a number of soil particles, forming an aggregate grain of considerable size. These soils in consequence generally have a somewhat • open structure, similar to that of a fine sand, through which water moves rapidly. When these soils are placed in a considerable quantity of water, as in the beaker method of elutriation, this cementing material dissolves sufflciently to break up the aggregate and release the smaller particles, thus materially changing the nature of the soil. This is most marked 173 174 METHOD OF MECHANICAL ANALYSIS OF ALKALI SOILS. in the case of the gypsum soils on account of the greater solubility of this cement. The solubility of such a cement is considerably increased by the presence of other salts in solution which have no ion in com- mon with those of the cementing material. The dissociated salts react with the calcium sulphate or calcium carbonate in solution to form, to a greater or less extent, all the chemical compounds possible through a rearrangement of the ions of the original substances. As a conse- quence, more of the cementing material goes into solution and the reactions continue until a condition of equilibrium has been reached. The equilibrium is, of course, disturbed by the addition of water or by changes in temperature. A more detailed statement of these mass reactions will be found in an accompanying paper by Dr. P. K. Cameron. From these considerations it becomes evident that neither the Osborne beaker method nor the elutriator method of Hilgard, both of which require large amounts of water, is applicable to the mechani- cal analysis of such soils. A trial of the beaker method on a gypsum soil from New Mexico confirmed these conclusions, fresh amounts of "clay" being liberated after each successive addition of distilled water. The determinations by this method differed widely, the amount of finer material increasing with the time employed in the separations. In fact, the amount of "clay" obtained when the soil was allowed to stand in contact with water for some time was sufficient to indicate a soil of close structure, permitting water to pass very slowly, instead of the open, porous structure which the soil was known to possess. It becomes evident that in order to get an analysis which would fairly represent field conditions it is necessary to make the analysis as rapidly as possible and with the use of a minimum amount of water. For this purpose the centrifugal method has been used with satisfactory results. The small amount of water required and the rapidity with which the separations can be efi'ected tend to reduce materially the amount of gypsum going into solution. The composition of the gyp- sum soil, as determined by this method, shows a much smaller quantity of clay and a proportionally larger amount of larger particles. This undoubtedly represents much more nearly the mechanical composition of the soil in situ. The mechanical composition of such soils must be, at best, somewhat indeterminate, owing to the fact that so large a percentage of the soil consists of water-soluble material. THE CENTRIFUGAL METHOD. The use of centrifugal force to hasten the deposition of particles in suspension naturally suggests itself as a means of shortening the tedious process of making separations by the method of sedimentation. Hop- kins' appears to have been the first to describe this method in detail, ' Proceedings of the Fifteenth Annual Convention of the Association of Official Agricultural Chemists, Bulletin 56, Division of Chemistry, U. S. Department of Agriculture, 1898, page 67. Report No. 64, U. S. Dept. Agr. PLATE XXVIll. O THE CENTRIFUGAL METHOD. 175 although the device had been previously used by several soil investi- gators. The method has been found so satisfactory in this laboratory on account of its quickness and convenience and its applicability to all soils that it is believed that a description of the apparatus and the method of operation will prove of interest. The apparatus necessary for making analyses by the centrifugal method is Inexpensive and simple in construction. The i)rimary requi- site is some means of securing the necessarily high velocity required for throwing down the soil particles from suspension. An electric motor is most suitable for this purpose if an electric-lighting current is avail- able. A water motor can be used, although subject to annoying inter- ruptions and fluctuations in speed unless ati independent water supply can be obtained. Some form of centrifugal apparatus operated by hand can be employed where but few samples are to be analyzed, although this method will be found to lack the advantage arising from an appa- ratus operated at a constant speed, such as may be obtained by the use of an electric motor. The centrifugal apparatus as designed by the writer for making mechanical analyses in the Division of Soils is illustrated in PI. XXVIII, The power is obtained from a Holtzer-Cabot, 110 volt, 16-inch fan motor. This motor uses a current a little in excess of that required for an ordinary 16-candlepower lamp, and will carry 4 centrifugal tubes of the dimensions described without serious heating. This style of motor is supplied with a rheostat in its base, enabling four different speeds to be obtained, which is a great advantage in making separations, besides enabling the motor to be gradually brought up to speed. The rheostat is also provided with an open contact point for stopping the motor. The fan and fan guard being removed, the motor is firmly screwed to a rigid supporting frame with its armature shaft vertical. A second hollow shaft, milled to fit the armature shaft, is slipped over the latter and fastened by a set screw. To the lower end of the hollow shaft are fastened four horizontal arms, each being about 8 cm. long and consist- ing of two parallel bars of 5 mm. square brass, 4J cm. apart A brass ring 5 mm. thick is trunnioned between each pair of bars at their free ends, and four light brass rods extend downward from this ring to a similar ring 15 cm. below. This trunnioned system swings outward and upward in the well-known way when the motor gathers speed. It is important that the system should swing freely, and care should be taken that the trunnion screws are sufficiently massive to stand the strain to which they are subjected at high velocities. Large heavy test tubes (18 by 3 cm.) serve admirably for the centrif- ugal tubes. The aperture in the upper metal ring is made large enough to admit the test tube easily, while the opening in the lower ring is somewhat smaller and provided with leather or cork washers, on which the test tube rests. A guard consisting of a screen of 5 mm. mesh 176 METHOD OF MECHANICAL ANALYSIS OF ALKALI SOILS. surrounds the movable portion of the apparatus as a safeguard against accidents. To protect the motor, the wires leading from the lighting current should contain fuses which will melt for currents exceeding two or three amperes. The analyses of four samples may readily be carried on at the same time. Ten grams constitute a suitable sample for analysis in an appa- ratus of the dimensions described. The preliminary preparation con- sists in agitating the sample of soil with about 200 cm. of water in a mechanical shaker' from six to eight hours, or until the surface of the larger grains, as seen under the microscope, appear to be clean and free from clay particles. A portion of the contents of each shaker bottle is transferred to its cor- responding centrifugal tube. The apparatus is then rotated for a length of time sufficient to throw down from suspension all particles larger than those which it is desired to retain in the finest separation. The "clay" water is then decanted into beakers and the remainder of the contents of the shaker bottles transferred to the tubes, the heavier material being thrown down as before. An important feature of the operation now claims consideration. In making additions of distilled water to the tube to effect further separations of "clay," it is desirable and important that this water should be forced in under considerable pressure. This forms the most satisfactory and convenient means of getting the material at thebottomof the tube into suspension again, being far superior to any agitation with a stirring rod or a rubber pestle, since it avoids all abrasion and the necessity of washing off the stirring rod each time. It will be found that thorough stirring of the material in the bottom of the tube by the jet of distilled water each time a decantation is made will materially shorten the time and diminish the amount of water rec^uired for an analysis. The apparatus for securing this pressure will be referred to later. When the "clay" has all been separated, as determined by a micro- scopic examination, using a micrometer, the tubes should be rotated for a shorter time, or at a lower rate of speed, leaving the particles constituting the next separation in suspension. Thewater containing these particles is then decanted into separate beakers and the process repeated until the separation of the second grade is effected. In making separations of particles exceeding 0.01 mm. in diameter, the sedimentation is sufficiently rapid to avoid the necessity of using centrifugal force. The distilled water is, therefore, added by means of the jet, and the material in suspension allowed to subside for a suita- ble length of time, as in the beaker method. Two separations, the clay (0.005 to 0.0001 mm.) and fine silt (0.01 to 0.005 mm.), are thus made by the use of centrifugal force. The silt (0.05 to 0.01 mm.), is separated by simple subsidence. The material remaining in the tube constitutes 'Whitney, Bulletin 4, Division of Soils, U. S. Department of Agriculture, 1895, page 9. THE CENTRIFUGAL METHOD. 177 the sands, which are dried and separated by means of sieves and bolt- ing cloth. The clay water does not usually exceed 600 cc, while the fine silt and silt together require about 500 cc. If these two last-named separa- tions are allowed to stand for a day or more, they will, of course, settle to. the bottom of the beakers, but the water in which they were sus- pended will be found somewhat turbid, indicating the presence of clay. No matter how carefully the separations may be made, this turbidity will nearly always occur, indicating a slight disintegration of these separations into finer material. This turbid water may be added to the water containing the "clay" in suspension if desired, although one will be justified in combining this suspended material with the sediment from which it was obtained. This latter method is preferable in soils containing large amounts of soluble material, such as the gypsum soils. If desired, the silt and fine silt sediments can be confined to very small volumes of water by again passing them through the centrifugal apparatus at a high velocity, which throws down all sedi- ments, leaving the clear water which may be decanted. This sediment may then be washed with a small quantity of water into small platinum evaporat- ing dishes. As recommended by Hopkins, it is highly desir- able to evaporate the whole of the clay water, the volume be- ing so small as to permit this being readily done. Porcelain piq. n, dishes are suitable for evap- orating the liquid to a small volume, when it may be transfeired to platinum dishes for ignition. In figure 11 is shown the device employed for obtaining a jet of distilled water under pressure. The air-tight reservoir R is filled by opening the valve A in the pipe leading from the distilled water tank, and by opening the vacuum cock V. After filling, V and A are closed and air admitted from the pressure cock P, thus submitting the water to a pressure of about 8 pounds per square inch. This gives a jet suf- ficiently strong to thoroughly stir up the sediment in the bottom of the centrifugal tubes, which through centrifugal action is packed so tightly in the tube as to permit an almost complete decantation of the water. The following mechanical analyses in duplicate serve to illustrate the accuracy which one may reasonably expect to attain in regular labo- ratory determinations, using the centrifugal method. These analyses were made by Mr. E. T. A. Burke, of this laboratory. H. Doc. 399 12 -Apparatus for water pressure in mechanical analysis. 178 METHOD OF MECHANICAL ANALYSIS OF ALKALI SOILS. Mechanical analyses in duplicate by ilie centrifugal method. Diameter. Millimeters. 2. to 1 1. to C 5 0.5 to 25 .25 to 1 .1 to 05 .05 to 01 .01 to 005 .005 to 0001 4333. Conventional uame. Fine gravel.... Coarse Band . . . Medium sand. . Fine sand Very iine sand. Silt Fine silt Clay LossatllOoC Loss on ignition i. Soluble salts Total 99.87 0.11 .66 7.78 42.67 27.04 .85 11.91 1.03 6.87 .95 0.09 .02 7.67 41.95 28.65 .94 11.80 1.80 .84 1.26 2.02 13.35 34.52 4.64 29.28 1.01 6.80 1.01 2.14 8.30 1.68 99.83 1.26 .88 1.16 2.40 13.38 35.49 4.52 28.37 2.18 8.16 1.69 TREATMENT OF SBPAEATIONS AFTER IGNITION. In making mechanical analyses it is customary to determine the organic matter and combined water by igniting to redness a separate five-gram sample, which has previously been carefully dried at 100° to 110° 0. It is consequently necessary to ignite also the several separa- tions of the mechanical analysis before weighing the sample. If calcium or magnesium carbonates are present they are partially reduced through ignition to the oxides of these metals. These oxides must therefore be converted back into carbonates before making the final weighing. The method commonly employed for this purpose, and the one recom- mended by the Association of Official Agricultural Chemists,^ consists in treating the ignited separations with a saturated solution of ammonium carbonate after being cooled. The reaction may be expressed thus: 0aO+(NH4)2CO3 % Ca0O3+2NH3+H2O The separations are then dried and afterwards heated to dull redness in order to drive off the ammonia, water, and excess of ammonium car- bonates. They are then cooled in a desiccator and weighed. Such treatment is not necessary for normal sodium or potassium carbonates, as these salts are stable and non- volatile at a red heat. This method answers very well if the separations contain only car- bonates. If, however, chlorids, sulphates, or nitrates are present two sources of error occur. In the first place, these salts are more or less volatile at the red heat required to burn off all the organic matter, so that an actual loss from volatilization will occur to some extent. Unfortunately, this difficulty is inherent in the method of determining organic matter by loss on com- bustion, which is the most satisfactory method available in connection with mechanical analyses. 'Methods of Analysis, 1898. Bui. No. 46, Div. of Chemistry, U. S. Dept. of Agr. TREATMENT OF SEPARATIONS AFTER IGNITION. 179 Secondly, if chlorids, nitrates, or sulphates are present in the ignited separations, the addition of ammonium carbonate to convert the oxides results in the formation of the ammonium salts of these acids. The reaction between the two salts is similar to that mentioned when speak- ing of gypsum in the presence of some other salt, and a condition of equilibrium is reached in exactly the same way. The class of reactions may be represented as follows : (1) 0aSO4+(NH4)i0O3 $ Ca0O3+(NH4)2SO4 (2) 2Na01+(NH4)2CO3 % Na20O3+2NH401 (3) 2N"aF03+(NH4)2CO,, ^ Na20O3+2]SrH4NO3. The ammonium salts of the above-mentioned acids are all volatile and will, of course, be driven off when the separation is heated to remove the excess of ammonium carbonate. The error from this source was called attention to some time ago by Hilgard and Jaffa.^ If chlo- rids, nitrates, or sulphates are present in considerable quantities, the above reactions will be indicated by the presence of the characteristic dense white fumes of the corresponding ammonium salts, which are freely given off at comparatively low temperatures. Ammonium car- bonate shows no fumes on vaporization. By successive treatments with ammonium carbonate it is possible to remove all of the negative or acid ions in this way, the positive or basic ions with which they were originally combined remaining as carbonates. The fact that a salt non-volatile at low temperatures is formed at the same time as the volatile ammonium salt reduces con- siderably the loss due to the above reaction, owing to the similarity in the reacting weights of the initial and residual salts as, for example, KaCl and J NaaOOs. The change in reacting weights for the three cases considered is given below : Salt. Eeacting weight. CaS04 68 60 68.5 53 85 53 I^aCl NaNOa NaoCOo This may, in part, account for the failure of soil analysts in general to consider this important source of error, particularly in connection with alkali soils. As pointed out by Hilgard and Jaffa,^ this loss may be avoided and the oxides converted into carbonates by treating directly with carbonic acid gas. A small amount of water is added to the ignited soil through which carbonic acid from a Kipp's apparatus is allowed to pass slowly for one or two hours. The separations are then evaporated to dryness and ignited at a dull red heat. The separate sample employed in determin- ing the organic matter must also be subjected to the same treatment. 1 Proceedings of the Society for tiie Promotion of Agricultural Science, 1888, p. 40. ^Loq (At. 180 METHOD OF MECHANICAL ANALYSIS OF ALKALI SOILS. DETERMINATION OF WATEE-SOLUBLE MATERIAL. A determination of the amount of water-soluble salts in a soil nas heretofore seldom been made in connection with a mechanical analysis. This determination is of so much importance and interest, especially in connection with soils requiring irrigation, that it is now performed in this laboratory for all soils. Two methods are available for this determination : First, the evap- oration to dryness of an aliquot part of the clear solution and weighing the solute; second, the determination of the concentration of the solu- tion through its electrical conductivity, computing from this data and the volume of the solution the amount of salt. The greatest difficulty in using the gravimetric method is to obtain the solution clear and free from suspended or colloidal clay. This is best effected by the use of repeated filtration through hardened filters. This process is, however, very laborious, and with some soils it is almost impossible to separate the clay, while with other soils the clay readily flocculates, either subsiding and leaving a clear supernatant liquid, or forming aggregates which can readily be removed by the filter. In most cases this flocculation appears to be due to a rather large amount of salts in solution. Evaporation to dryness will some- times prevent clay from becoming suspended when water is again added. Ignition will nearly always accomplish this, but at the expense of a portion of the salt content, as will be shown later. The electrical method of determining the soluble salt content has the important advantage of not requiring any filtration of the solution, the presence of the suspended clay having no appreciable influence upon the conductivity. The determination is consequently very rapid, consisting simply in measuring the specific resistance of the solution and noting its temperature and volume. The resistance of the solution is conveniently determined with the apparatus previously described by the writer.' The transformation of the observed electrical resistance of the solu- tion into terms of solution concentration may be performed by one of the three following methods: (1) The approximate composition of the salts in an area throughout which the composition (not the amount or concentration) may be fairly assumed to be uniform is determined by chemical analysis. An aqueous solution of these salts of known concentration is then prepared, the salts being present in the solution in the proportions previously found by the analysis. From this standard solution other solutions of regu- larly diminishing concentration are prepared and the electrical resist- ance of each determined in the electrolytic cell to be used in subse- quent measurements. From these data a resistance-concentration curve ' Electrical Instruments for Determining the Moisture, Temperature, and Soluble Salt Content of Soils, Bulletin No. 15, Division of Soils, U. S. Department of Agri- culture, 1899. DETERMINATION OP WATER-SOLUBLE MATERIAL. 181 can be constructed, from which the concentration of any unknown soil from that area can at once be determined if its electrical resistance is known. The original solution must exceed in concentration any of the solutions of unknown concentration, while all solutions must be reduced to a common temperature. This reduction is very simple, a table for this purpose being given in Bulletin No. 7 of this division. The weight of the soil used and the volume of the solution being known, the amount of the soluble salts in any sample can be expressed as a percentage of the dry weight of the soil. As far as accuracy is concerned, this method is practically as reliable as a direct determina- tion by the gravimetric method. On account of the time required for the necessary chemical analysis this method is not well adapted to a determination of the salt content of a small number of samples unless a chemical analysis of the water-soluble material has been made for other purposes. (2) The second method of determining the soluble salt content con- sists in making a concentrated solution from a soil crust or a soil highly charged with salts. A portion of this solution is evaporated to dry- ness, in order to determine the concentration, while from the remaining portion a series of solutions of decreasing concentration is formed, the resistance of each solution being determined as before and the resist- ance-concentration curve prepared. This method requires much less preliminary work than the first, and is just as accurate, having a gravi- metric basis. A quantitative knowledge of the composition of the sol- uble material in this case is not obtained. (3) If the sample is an isolated one, or if the number of samples from any locality is so small that it is not desired to prepare a resistance- concentration curve, a very fair approximation to the salt content can be obtained by using a resistance-concentration curve for some single salt, such as sodium chlorid. For dilute solutions, all of the more strongly dissociating salts, such as chlorids, nitrates, and sulphates, are practically ionized to the same extent. Aside from the slight differ- ences in the migration velocities of the ions, which in this case may be neglected, these strongly dissociating salts in dilute solution have practically identical concentration-resistance curves, if we use ionic concentration in place of the conventional mass of the solute, or, in other words, if we express the concentration in gram-molecules instead of in percentage of solute by weight in the solution. Therefore, it is practicable to use this common concentration-resistance curve (the con- centration being expressed in gram-molecules) to determine the salt content of dilute soil solutions. To convert this determination Into grams of solute per liter, it is simply necessary to multiply it by the reacting weight of the salt in the terms of which it is desired to express the concentration. The per cent of salt as determined by this method will consequently vary somewhat, depending upon the reacting weight of the salt chosen as a basis of computation. For example, if the 182 METHOD OF MECHANICAL ANALYSIS OF ALKALI SOILS. solute was computed as sodium chlorid the factor used would be 57, while for sodium sulphate (a bivalent salt) the factor would be 142/2, or 71. Using a coucentration-resistance curve computed on a gram -molecular basis, we can thus compute the salt content in terms of any salt desired. If it is preferred to express all results in terms of some single salt or a combination of salts, a resistance-concentration curve can be prepared directly as in the first two methods, the concentration being expressed in terms of the mass of solute. In the salt determinations made in this laboratory in connection with the mechanical analyses no effort has been made to maintain a constant relation between the amount of soil and the volume of the solvent. The volume of the water necessary in separating the clay (about 600 cc.) is so large in comparison with the amount of salt in a 10-gram sample of soil that a considerable variation in the volume of the solvent would have but little influence upon the amount of salts in solution, except in the case of slightly soluble salts. Since these salts under soil con- ditions exert but little influence upon the plant, so far as their concen- tration is concerned, the procedure outlined appears justifiable. The "clay" water is generally found to contain the principal portion of the soluble matter, which is determined and subtracted from the weight of the combined clay and salt separation to give the amount of clay. The presence of any appreciable amount of salt in the "fine silt" or "silt" waters is easily determined by the Instrument; audit can be neglected, combined with the clay water, or determined and subtracted from the "fine silt" or "silt" determinations as the case demands. The following table gives the salt content of three alkali soils deter- mined by three different methods. These soils were selected to show wide variations in their soluble salt content. The determinations were made by Mr. W. G. Smith of this laboratory. Salt content. Catalogue No. 1. 2. 3. 4U5 Per cent. 1.60 5.40 3 Per cent. 1.08 4.65 2.70 Per cent. 4119 4 55 4131 --- --- 2 63 3.33 2.81 2 71 The first column of the table gives the catalogue number of the soils investigated. Column 1 gives the per cent of salt found by the elec- trical method after two grams had been left to stand for six days in contact with 600 cc. of water. The salt is expressed as the percentage of the dry weight of the soil. Column 2 gives the per cent of salt found after evaporating the solutions used in 'So. 1 to dryness, igniting, dilut- ing to 600 cc, and allowing them to stand for six days. Column 3 gives DETERMINATION OF WATER-SOLUBLE MATERIAL. 183 the salt coutent determined gravinietrically by evaporating to dryness the clear solution used for the determination given in column 2, No difficulty is experienced in obtaining a clear solution from any ignited soil, the clay seldom showing any tendency to remain in suspension after ignition. The results of the table show that the determination of the salt con- tent after ignition is not advisable, a considerable portion of the soluble matter either being driven off during ignition or else converted into an insoluble form. The results of columns 2 and 3 show a satisfactory agreement between the electrical and gravimetric determinations, con- sidering that the concentration-resistance curve used in converting the electrical determinations was one prepared for a series of soils, of which the three samples selected formed but a small proportion. The salt determinations in the two duplicate analyses given on page 178 serve to illustrate further the accuracy of the electrical method. SALTS AS INFLUENCING THE RATE OF EVAPORATION OF WATER FROM SOILS. By LYMAN J. BRIGGS. INTEODUCTION. The physical properties of water which iufluence the rate of evapo- ration from a soil surface are surface tension, viscosity, and vapor pressure. Since these factors are all modified to a greater or less degree by the addition of different quantities of salts, it is of interest to consider the iniiuence of various salts in solution upon the evaporation of water from soils. The surface tension of an aqueous solution of a salt is higher than that of the solvent. It appears from the work of Dorsey' upon dilute solutions and of Sentis^ upon solutions greater than one-half normal that the surface tension is a linear function of the concentration. This may be represeuted by in which T^ is the surface tension of a solution containing e gram equiv- alents per liter, T,v the surface tension of water at that temperature, and iL a constant varying with different salts. For sodium chlorid this constant K may be taken as 1.5 dynes, and for sodium carbonate as 2 dynes per square centimeter. Since the surface tension of pure water is about 73 dynes per square centimeter at 18° C, a difference of one gram equivalent in the concentrations of two portions of the soil solution would cause an increase in the surface tension of 2 per cent for sodium 'jhlorid and 2.6 per cent for sodium carbonate. The writer' has previously pointed ouifc that the capillary movement of water in soils is due to the pressure of the peculiar capillary surface formed by the water in the capillary space about the point of contact 'Physical Review, 5, p. 228, 1897. »Jour. de Ptys. (3), 6, p. 183, 1897. 3 The Mechanics of Soil Moisture, Bulletin No. 10, Division of Soils, U. S. Depart- ment of Agriculture, 1897. 184 INTRODUCTION. 185 of two soil grains. The pressure of such a surface may be repre- sented by iu which P is the pressure, T the surface tension, and rx and r^ the two principal radii of curvature. Since T is involved only as a linear func- tion, a small change in T can only produce a proportional change in P. Consequently the small changes in T, which amount to only 2 per cent in the extreme variations considered, can practically be neglected as compared with the much greater variations in the value of r. In other words, the salts in solution play a minor part as compared with the curvature of the water surfaces in determining the capillary movement of water in soils. Since the viscosity of a solution increases with the concentration, the presence of large amounts of soluble salts would retard to some extent the movement of water in the soil, and so lessen the rate of evaporation by limiting the movement toward the surface. Arrhenius has found that the relation between viscosity and concentration for a large number of solutions can be expressed by the equation: in which ju^ is the specific viscosity of a normal solution referred to the solvent at the same temperature and n the number of gram-molecules in the solution under consideration. For example, a normal solution of sodium chlorid has a specific viscosity of 1.097, or about 10 per cent greater than water. The specific viscosity of a two-normal solution would be 1.20 and a half-normal 1.04. Since a saturated solution of sodium chlorid at ordinary temperatures contains approximately five gram-molecules per liter, the specific viscosity under such conditions would be 1.59. In alkali soils, therefore, where the salts are present in excess, the increased viscosity of the solution would decrease the rate of movement of water to a considerable extent. It should be noted in this connection that the effect of increased concentration is to increase the surface tension, and thus increase the rate of movement. The two factors, consequently, oppose each other, and the resultant change in the rate of movement of water due to change in concentration is much less than if either factor existed alone. Finally, the vapor pressure of a salt solution decreases with increase of concentration. The vapor pressure of a normal sodium sulphate solu- tion at 100° t). is 25 millimeters less than the pressure of water vapor at that temperature. The variation in vapor pressure from that of the solvent at other temperatures has not been determined for this salt; but approximate values can be ascertained by comparison with the approximate normal solution of potassium hydrate, for which the vapor 186 SALTS AS INFLUENCING EVAPORATION FROM SOILS. tension at various temperatures has been determined.' In the follow- ing table the first column gives the temperature, the second and third columns give the vapor pressure of the solution and of the water at the temperature considered, while the fourth column gives the temperature to wliich it would be necessary to cool pure water in order to make its vapor pressure equal to that of the potassium hydrate solution : Tem- Vapor pressure. Tempera- ture of water for perature 16. 6 per cent Water. equal 0°C. KOH. yap. pr. 10 8.01 9.14 8.1 16 11.86 13.51 13.9 20 15.25 17.36 17.9 25 20.67 23.52 22.8 30 27.74 31.51 27.8 34 34.84 39.52 31.8 It will be seen from the table that the evaporation from a normal salt solution at 30° C. into vapor free space is the same as with pure water at 27.8°. A certain conservation of the moisture would conse- quently follow. However, even with a normal solution this is not very marked and could only have an important effect in case the soil mois- ture is saturated with the salt. A saturated solution of sodium car- bonate would correspond in rate of evaporation to a lowering of temperature of about 10° 0., assuming the evaporation to be taking place in a vacuum. There is another factor which probably exerts more influence upon the rate of the evaporation of water than any of the factors mentioned, namely, the physical condition of the surface of the soil. Hilgard has pointed out that some of the so-called alkali salts, especially sodium carbonate, exert a marked influence upon the physical character of the soil. Buffum^ calls attention to this fact and also states that alkali soils may appear damp when adjoining lands seem dry. To explain this phenomenon he conducted experiments on soils which had been moistened with solutions of salts. Thesalts used were sodium carbonate, sodium chlorid, sodium sulphate, and magnesium sulphate in solutions varying in concentration from 0.1 per cent to 9 per cent. The general results of the experiments show that the amount of water lost from the pots containing 9 per cent solutions (which is equivalent to 2.2 per cent alkali, in a soil containing 25 per cent of the solution) was approximately one-half as much as that from soils containing the same percentage of pure water. The effect of the presence of soluble salts in the soil on the evaporation, in Buffum's experiments, was so much 'Landolt and Bornstein, "Phys. Cheni. Tab.," S. 68. 'Bulletin No. 39, Wyoming Experiment Station, p. 55, 1898. SODIUM CHLORID. 187 greater than would be expected from the theoretical considerations given in the first part of this paper that the writer was led to believe that this difference was due to a change in the surface conditions of the soil from the accumulation of salts rather than from the presence of the salts in solution. In other words, either the formation of a mulch or the crusting of the soil at the surface had taken place, preventing evaporation to some extent. To test the validity of this opinion the following experiments were planned : (1) The evaporation of water from soils spread out in thin layers in large trays, one sample being moistened with distilled water and the other with a salt solution. The soils in both trays to be stirred from time to time in order to prevent any possible crusting of the soil. This would indicate whether the abnormal lessening of the evaporation observed by Bufifum was due to a surface condition. (2) The relative rate of evaporation in open dishes of pure water and solutions of the concentration used in experiment (1), as a check upon the conditions observed during that experiment. (3) Evaporation from the surface of soils packed in tall, narrow cylinders and moistened with pure water and salt solutions, as a repeti- tion of Buffum's experiments. This arrangement would serve to accentuate the effect of the crusting of the surface. The soil used in all the experiments was from James Island, South Carolina, being the soil upon which the Sea Island cotton is grown. The texture is shown in the accompanying mechanical analysis, made by Mr. W. Gr. Smith, of this laboratory : Diameter. Miilimetert. 2 too. 1 1 to .5 .5 to .25 .25 to .1 .1 to .05 .05 to .01 .01 to .005 .005 to .0001 Conventional name. Fine gravel. ... Coarse sand — Mediiim sand.. Fine sand Very fine sand. Silt Fine silt Clay Loss at 110° C... Loss on ignition . 80. James Is- land, S. C. Per cent. 0.00 .37 6.01 79.15 2.50 .65 4.23 4.60 1.85 SODIUM CHLORID. Experiment 1. — To test the influence of the salt content of the soil uj)on the rate of evaporation of water from the soil, two samples of soil containing 200 grams each were moistened, respectively, with 40 grams of pure water and normal sodium chlorid solution. These soils were spread out in two shallow photographic trays of the same size, meas- uring approximately 8 by 10 inches. The soil containing the salt solu- tion was frequently stirred, in order to prevent the possible formation of a crust at the surface. The soil in the other tray was also stirred at 188 SALTS AS INFLUENCING EVAPORATION FROM SOILS. the same time, in order to treat the two soils as nearly alike as possible. The results of the experiment are given in the following table, and shown graphically by the accompanying curve (fig. 12): Evaporation from soil exposed in iraya, moistened with water and normal XaCt aolution. Time (min- utea). Water evaporareil (grams). "Water. NaCls> lutiuu. 50 90 130 160 235 295 11.5 20.6 29.8 32.3 36.2 39.5 10 18.8 20 28.5 32.1 35.5 As it was not possible to keep the conditions of evaporation constant, the time intervals serve only as an indication of the rate at which the -^ ' — ' 4^ ij^ , ^ ^^ -- / ^A "Kal kso 'M^ yn A / /■ V / 40 35 i 30 I 25 zo ;5 ^ w Time — hours. Fig. 12 — Diagram showing evaporation from soil in trays, moistened with water and NaCl solution. evaporation was taking place. The drying was carried on in a room at a temperature of about 25° 0., the air above the trays being kept constantly in motion by means of a fan, and the trays being frequently interchanged to prevent an error due to their position. It is seen in the table that the rate of evaporation was about 10 per cent greater from the soil moistened with distilled water. It should be remembered that the concentration of the solution in the soil containing the salt constantly increases as the evaporation goes on, and on this account the evaporation should gradually decrease. The actual concen- tration of the solution in the soil is a matter of more or less conjecture, since the measurements of the electrical conductivity of a soil mois- tened with salt solution indicate that an appreciable amount of the salt originally added in the form of solution does not remain in solution in the presence of the soil grains. The conductivity of a solution is at least very materially lowered by the addition of the soil, indicating that SODIUM CHLOEID. 189 physical absorption has taken place and that part of the ions of the salt are no longer free to move. Experiment 2. — To determine whether the difference in the evaporation observed in Experiment 1 was more than could be accounted for by the difference in the physical properties of the salt solution and pure water, an experiment was conducted to determine the difference in the rate of the evaporation of the two solutions. One hundred grams of distilled water and of a normal solution of sodium chlorid were placed in two Petri dishes 3| inches in diameter and iiveeighths of an inch deep. These dishes were so placed as to be subjected to the same conditions, although no attempt was made to keep these conditions constant. The rate of evaporation is shown in the accompanying table: Table of maporation from Petri dishes Distilled NaCl solu- Time. water, total tion, total loss. loss. Hours. Grams. Grams. 1 .9 .9 4 2 2 23 12.8 11.8 27 15.3 14.3 30 18.3 17.1 49,5, 35.3 32.3 53| 41 37.7 95 i 63.5 .IS. 1 121H 74.3 66.8 1435' 82.6 73.3 E 65 M 50 ' I From this table it will be seen that the relative rate of evaporation of the distilled water and the salt solution was not sensibly different when freely exposed in evaporating dishes than when mixed with the soil. From this we are justified in concluding that the evaporation of water from the salt solution is not necessarily modified to any appreciable extent by havingthis solution in con- tact with the soil grains. In other words, the evap- oration of the water from a soil would be modified only by the presence of soluble salts so far as these soluble salts would influ- ence the evaporation of a solution of similar concen- tration when freely ex- posed to the air. This would not amount, as we have seen, even in the case of normal solu- tions, to more than 10 per cent. This is far from sufficient to explain the great difference noted by Buffum. ■ /• y / / / ,' ■■ / ,' 'A'^' y -• * .4' y y I /.' ' ^ // / /; / / / iO ZO 30 4>0 JO 60 7a 80 &0 JOG 110 1Z0 130 140 150 Time— hours. Fig. 13.— Diagram showing evaporation from Petri dishes, distilled water, aud normal sodium chlorid solution. 190 SALTS AS INFLUENCING EVAPORATION FROM SOILS. Experiment 3. — It remained to test the opinion of the writer — that the results obtained by Buffum were due to the formation of the crust at the surface of the soil. Accordingly two cylinders 2 inches in diameter and 11 inches high, holding 800 grams of soil, were filled with samples of soil moistened with 20 per cent of distilled water and with normal sodium chlorid, respectively. To supplement these, a third cylinder of the same dimensions was filled with a soil containing 20 per cent of distilled water to within one-eighth of an inch from the top of the cylinder. The remaining one-eighth inch was filled with pulverized sodium chlorid. These cylin- ders were kept under similar conditions and the loss of moisture determined by weighing them from time to time. The results are given in the accompanying table: Table of evaporation from cyHiidera of soil containing SO per cent distilled vjater, : cent normal NaCl, and NaCl on surface. I per Time. Distilled water, total loss. Normal NaCl Bolntion, total loss. NaCl on surface, total loss. Hours. 17 24 65 120 148 170 197 222 245 288 337 Grams. 6 9 19 27 32 38.5 46 53 60 70 81.5 Grams. 0.6 3.5 9 13 15.5 18 21 24.5 27.5 30.7 35.5 Grams. 6 10.5 13.5 15 17.5 21.5 24 28.3 33.5 The rate of evaporation of water in the three tubes is more clearly shown in the accompanying diagram. The irregularity of the curves is due to the fact that uniform conditions of evaporation could not be maintained. It will be noticed, however, that these irregularities appear in all curves, showing that the tubes were similarly influenced. From these curves it is evident that the rate of evaporation from the soil containing distilled water was about twice that from the soil moistened with the normal salt solution, confirming in a general way Buffum's experiment. Consequently, we are justified in concluding that for this salt the marked diminution in evaporation is due mainly to surface conditions. A peculiar condition was noticed in connection with the soil which was moistened with the salt solution. During the evaporation of the water the salt crystallized just beneath the surface of the soil, leaving it in a very loose and porous condition; in fact, after the evaporation had proceeded some time, the surface of the soil in this tube was raised one-half inch or more above the mouth of the tube, due to the forma- tion of minute crystals of the salt interspersed among the soil grains. This tends to confirm the idea that the decreased evaporation in an SODIUM SULPHATE AND SODIUM CARBONATE. 191 alkali soil is due to the formation of a natural mulch, through the crystallization of the soluble salts at the surface, which so loosens the surface as to decrease the evaporation materially. In the tube in which the soil was covered with a layer of pulverized salt no loosening of the surface was noticed, but the salt seemed to form a solid incrustation over the surface of the soil, reducing the evaporation. The equality in the slope of the two salt curves is impor- tant in showing that the influence of the concentration on viscosity and surface tension need not be considered in this case. SODIUM SULPHATE AND SODIUM CAEBONATB. whether as 80 tlie conclusions drawn from the es eo 46 ss In order to determine experiments with sodi- um chlorid applied to all of the more abundant "alkali" salts, experi- ments similar to those just described were made with sodium sulphate and sodium carbonate by Mr. M. H. Lapham, of this laboratory. Experiment 1. — T his was conducted practically the same as described in the corresponding experi- ment with sodium chlorid, except that the determi- nations for each salt were made in duplicate. Pour 200-gram samples were exposed in trays, two be- ing moistened with 40 grams of water, and the others with 40 grams of a normal solution of the salt under investigation. The results given in the following table are a mean of the duplicate determinations. These are also expressed graphically in figures 15 and 16. It will be noticed that at first the salt has little influence upon the rate of evaporation of the water from the surface of the soil. As the concentration increases and the soil becomes drier the influence of the salt is more marked. In this experiment the. greatest diminution in the rate of evaporation produced by the sodium sulphate does not exceed 10 per cent. The experiment with sodium carbonate (figure 16) gave practically the same results. f / / / / / / X Y ^/ ;^ f / ^,. ■ y ^ / / n?'-' ,-•'' y' / / •A 9^i .-0,0 t^' / / t ^9- i\ f / ,— --' ■ ''' • ■'■H r / /'' .-•' ^ .^" // 30 60 SO ISO 150 JSO Z10 340 S70 300 330 360 Time— hours, Fig. 14.— Diagram showiug evaporatioD from cylinders mois- tened with distilled water and sodium chlorid solution. 192 SALTS AS INFLUENCING EVA.POBATION FROM SOILS. Taile of evaporation of distilled water mid normal Ka„SOifrom soil in trays. Distilled Normal Time. water, total Na.jSOa, to- loss. tal loss. Hours. Orams. Qrams. Oi 1 1.30 1 2.40 2.55 li 3.60 3.65 2 5.55 5.60 24 7.50 7.30 3 9.70 9.30 3i 11.55 10.90 4 13.30 12.45 *J 14.85 13.85 5 16.70 15. 55 5i 18.40 16.90 6 20.30 18.50 6i 22.15 19.90 7 23. 55 21.15 7i 25.25 22.75 8 26.90 24.25 8i 28.20 25.45 9 29.45 26.60 9* 30.65 27.70 10 31.70 28.55 32 30 3Ji a? 2? zo • /* ys ' /-t /2 m s e 4 Z / / / / / /, / i/ r / i 1 ' ■ '''<^' o>, Oii ) / / / 1 ft ,/ < / / ( / / / O 2 J2. 3 ^ S 6 7 $ 9 /O Time — hours. !Fia, ]5. — Diagram showing evaporation from soil in trays moistened with distilled water and sodium sul- phate solution. SODIUM SULPHATE AND SODIUM CARBONATE. 193 lahle of evaporation of distilled water and normal Na^COg solution from soil in trays. Distilled Normal Time. water, total NajCOs, loss. total loas. Hours. Grams. Orams. Oi 2.25 2.15 1 3.95 3.70 li 5.80 5.35 2 8.30 7.80 2i 9.85 9.30 3 11.30 10.50 H 13.50 12.45 4 15.30 13.90 ^ 17.35 15.75 5 19.40 17.65 5i 21.30 19.20 6 22.60 20.25 6i 24.05 21.35 7 25.05 22.15 7i 26.10 22.95 8 27.05 23.90 8i 28.25 24.80 9 28190 25.55 94 29.55 26.05 10 30.30 26.65 32 SO ss 26 24, £2 SO 18 16 ■14 1Z 10 8 e 4, Z 01S34S6783 10 Time — hours. Fia. 16.— Diagram sliowing evaporation from soil iu trays moistened with distilled water and normal sodium car- bonate solution. Experiment 2 A. — These experiments were conducted for the piu'pose of determining the relative rate of evaporation from the surfaces of pure water and salt solutions under ordinary atmospheric conditions of temperature, pressure, and humidity. The water and the salt solu- tions were exposed in Petri dishes of the same diameter and depth. H. Doc. 399 ^13 / ■y' / / , / / • 4' / f. / / / > i/i ^. 4 w If t //' / / / / 11)4 SALTS AS INFLUENCING EVAPORATION FROM SOILS. The air above the dishes was kept gently moving by means of an elec- tric fan, and all i>recautions to obtain uniform conditions were observed. The results of the experiments are given in the following table and shown graphically in flg. 17 ■ Table of evaporation of diaUUed water and normal Na^SOt and NafiOs solutions from Petri dishes. Distilled Normal Normal Time. water, total Na™S04, to- tal loss. NajCOa, to- loss. tal loss. Sours. Grams. Orams. Orams. 1 1.10 1.05 1.10 2 2.20 2.10 2,20 3 3.45 3.20 3.30 4 4.45 4.30 4.40 5 5.15 5.00 5.10 6 6.00 6. SO 5.80 7 6.75 6.50 6.60 8 7.50 7.30 7.30 9 8.40 8.20 8.20 10 9.40 9.10 9.20 11 10.35 10.00 10.10 12 11.25 10.80 10.90 13 12.30 11.80 11.85 14 13.35 12.80 12.80 15 14.15 13.60 13.70 ■ 16 15.10 14.50 14.60 17 15.90 15.20 15.30 18 16.75 16,00 16.20 19 17.60 16,90 17.00 20 18.15 17.40 17.50 21 18.80 18.00 18.20 22 19.50 18.80 18.90 23 20.25 19.40 19.50 24 21.05 20.20 20.30 25 21.60 20.70 20.90 26 22.40 21.40 21.60 27 22.95 22.00 22.20 28 23.50 22.50 22.70 29 24.10 23.00 23.20 30 24.65 23.50 23.70 31 25.25 24.00 24.30 32 25.96 24.60 24.90 33 26.50 25.10 25.50 34 27.10 25.60 25.90 The evaporation curves for the solutions are displaced to the right to avoid confusion. The concentration beginning with normal solu- tion increases with evaporation until a saturated solution of sodium sulphate is obtained. The three solutions, however, show very little difference. The sodium carbonate solution evaporated a little more rapidly than the other, probably due to the fact that it does not dis- sociate so strongly as sodium sulphate, and hence has a higher vapor pressure. Neither of these solutions show anything like the diminution in the rate of evaporation which we should expect from theoretical con- siderations of the lowering of the vapor pressure due to the concentra- tion. Consequently, we are led to conclude that the rate of the diffu- sion of the water vapor under ordinary conditions is the controlling factor, and that the concentration of the solution would play the more important part only when the surface of the solution is continually being supplied with vapor-free air. Uxperiment 2 B. — In order to substantiate more completely the con- clusion that the evaporation from reservoir or soil is not diminished bv SODIUM SULPHATE AND SODIUM CARBONATE. 195 the presence of salt in solution to the extent which consideration of the lowering of the vapor pressure would lead us to assume, a determina- tion was made of the relative rate of evaporation of water and of satu- rated solutions of the three salts previously used. The solutions were exposed in Petri dishes as ^s before, while the surfaces ze of the liquids were con- ^^ stantly swept by a strong ^^ blast of air from an elec- tric buzz fan. The blast of . *" air was sufficiently strong | '* to produce pronounced ^ ie ripples on the surfaces of the liquids, particularly that of the distilled water, the increased viscosity of the solutions somewhat diminishing this effect. Therefore, the conditions of this experiment — the rapid movement of air over saturated solutions — are such as to give the maximum influence of the salts investigated on the diminution of evaporation under the given conditions of temperature and pressure. The results are shown in the following table and curve, fig. 18 : Table of evaporation of distilled water and saturated solutions of NaCl, Na„SO.i and Na-iCOsfrom Petri dishes. S '-* N / / •" / / /' f/ ^ / / ■^ f / /' f. /■ 4 o> '^^* / # ^ J / / T' / /" / / / > / / /' / / / / z o * a t/t^so^ o * f/a^ Cffj 8 ie 20 J2 zs 36 ss 18 36 3S Fig. 17.- 16 11 B 12 Time — hours. Diagram showing evaporation from Petri dishes, normal solutions. Time. Dist. water, NaCl, NasSO,, Na^COa, total loss. total loss. total loss. total loss. Hours. Grams. Grams. Grams. Grams. i 1 0.8 0.9 1 1 2.3 1.8 .2.1 2.3 li 4.2 3.5 4.0 3.9 2 5.6 4.7 6,5 5.3 2i 7.8 6.4 7.5 7.3 3 9.7 8 9.4 9 3i 11.1 9.3 10.7 10.2 i 12.9 11 12.6 11.8 4i 14.8 13.6 14.3 13.2 5 16.5 14.2 15.9 14.4 Si 17.8 15.2 17.3 15.5 6 19.1 16.4 18.6 Crystallized 6i 20.5 17.6 19.9 7 Crysts lUzed. These results show a maximum diminution in the rate of evaporation of about 14 per cent. The average temperature of the laboratory was 19° C, while the average relative humidity was 45 per cent. An aver- age of 15 grams of water was evaporated from each dish during six hours. Since the diameter of the dish was about 8.4 cm., this would correspond to an evaporation of 1.1 cm. per square centimeter of sur- face or 0.44 inch per square inch during twenty-four hours. The humidity and rate of evaporation are consequently such as to make the results of this experiment applicable to arid conditions. 196 SALTS AS INFLUENCING EVAPORATION FROM SOILS. Experiment 3. — This experiment was conducted similar to that for sodium chlorid. Three cylinders were filled with soil moistened with £2 20 /€ 16 i '^ I I 10 / / / / */ / W m> '<* /' , e )/ 1 A* \A f / / 4?'' / / / I f y / / // / / / ^' / /* r 2 a 4 .5 n 7 8 ,9 N/iz SO^ I «? a 4 .■) e> 7 (S m^co^o / 2 3 4 3 6 7 Time — liours. Fig. 18 — Diagram showing evaporation from Petri diabes, saturated solutions. 20 per cent by weight of distilled water and normal solutions of sodium sulhpate and sodium carbonate, respectively. In a fourth cylinder the soil was moistened with distilled water and mulched with a layer of sodium sulphate one-eighth inch thick. The results are shown in the accompanying table and curves, fig. 19: Table shotoing evaporation from soil cylinders — Disiilled water, normal Na^ CO3 and Nai > -' i/ ..0? / ^' % f f/ ■fO* / / i^" / 1 > // / /, \ / less marked, on account of the remarkable degree to which this salt is absorbed by the soil and kept from accumulating at the sur- face. The crystallizing out of the soil at the surface does take place to some J extent, but the white crys- -i ^° tals are masked by the ac- | companying organic mat- £ ter which this salt bripgs | into solution. The result- ^ ant dark color of the sur- face, which so characterizes the presence of this salt in alkali soils, was very marked in the sodium car- bonate tube during the first part of the experiment. This is shown to some ex- ten t in Plate XXIX. The tube containingthe sodium sulphate as a mulch does not differ materially in its rate of evaporation from the tubes containing the solutions, showing that the conditions within the soil are secondary to surface conditions in alkali soils. Sodium carbonate does not form at the surface the loose, fluffy mix- ture of soil grains and salt crystals so characteristic of sodium sul- phate and sodium chlorid, but tends rather to the formation of a crust which is more or less broken and tilted by the subsequent crystalliza- tion beneath. This action, of course, depends upon the texture of the soil, and in field conditions this would be influenced by the presence of other salts. s a 9 10 * 5 6 Time— days. riQ. 19.— Diagram showing evaporation from cylinders of soil moistened witli distilled water and with normal so- dium carbonate and sodium sulphate solutions. 198 SALTS AS INFLUENCING EVAPORATION FROM SOILS. CONCLUSIONS. From the results given ia the paper the following couclusions may be drawn : (1 ) Salts influence the evaporation of water from the surface of a soil by changing the surface tension, the viscosity and the vapor pressure ot the system, and the physical character of the soil, particularly at the surface. (2) The surface tension and viscosity influence the rate of evapora- tion only through the modification of the rate of capillary movement. Both surface tension and viscosity increase with increase in concentra- tion. An increase in surface tension increases the rate of capillary movement, while increase in viscosity diminishes it. These two factors, consequently, oppose one another. (3) The rate of capillary movement within a soil is of secondary importance to the physical character of the surface, as modified by the presence of crystallized salts. (4) The relative rate of evaporation from a soil moistened with pare water and salt solution, respectively, is the same as for the water and solution without the soil, providing no surface mulch is formed. (5) The rate of evaporation gradually decreases with increase in con- centration. (6) The diminution of evaporation with increasing concentration is much less than the corresponding diminution in vapor pressure. This is due to the fact that the atmosphere is never vapor-free and that the diffusion of water vapor is retarded by the surrounding air. (7) The diminution of evaporation of soils containing solutions of "alkali" salts is much greater than can be accounted for through the influence of the lower vapor pressure, and is due to the formation of a mulch at the surtace of the soil through the crystallization of the salts. O w&m^^ ■:ViiWm»if:K:fm^;''. .»7W' :^f-.- ;?>^ ^••Wwmm :U 'lf,V'- .("".T. '^X^(^' >^i^j^Ji£ U>:m ";«/■ ".iVlV i ilV 7.*.T ■IV,/, '.4i If-, b'.% ■'.'.'■./I' 1 ■ 1- .'. * v.' i ...■•y'l-..'?-'** ^•: <^-'^i^'f5:i?^iV:.^Af ■^■^"V*r''^-? t- V'^5(«g!V .V