> H ^ 
 
 
 
 
 
 
 
 
 ^> V 
 
 
 
 
 
ASIATIC AFFINITIES 
 
 THE OLD ITALIANS. 
 
 KOBERT ELLIS, B.D., 
 
 FELLOW OF ST. JOHN'S COLLEGE, CAMBRIDGE, AND AUTHOR OF 'ANCIENT 
 ROUTES BETWEEN ITALY AND GAUL." 
 
 Tuseos Asia sibi vindicat.' 
 
 LONDON : 
 
 TRUBNEE and Co., 60, PATERNOSTER BOW, 
 
 1870. 
 
.£4- 
 
 T. RICHARDS, PRINTER, 37, GREAT QUEEN STREET. 
 
 
CONTENTS. 
 
 CHAPTEE I. 
 
 PAGE 
 
 Migrations of the Aryans from their original home in Mount 
 Imaus— Eoute of the Thracians into Europe — Western 
 and Northern limits of the Thracian area in Europe — 
 With what nations the Thracians came into collision — 
 Primitive population of the South of Europe, and of Asia 
 Minor and Armenia 1 
 
 CHAPTEE II. 
 
 Etruscan sepulchral inscriptions 26 
 
 CHAPTEE III. 
 
 Etruscan votive inscriptions 78 
 
 CHAPTEE IV. 
 
 The Inscription of Cervetri 95 
 
 CHAPTEE V. 
 Conj ligations and Numerals 114 
 
 Index of Etruscan words, with their meaning 153 
 
ERRATA. 
 
 In p. 15, 1. 17; p. 16, 1. 1; and p. 18, 1. 24; for Mnesia read Moesia. 
 
 I n p. 94, 1. 4 ; for sulthn read salthn. 
 
 In p. 102, 1. 27; for mataram read mataram. 
 
 i 
 
PREFACE. 
 
 The publication of the great work of Fabretti, the 
 Corpus Inscriptionum Italicarum, has rendered avail- 
 able a rich store of materials for investigating the 
 affinities of the ancient languages of Italy. The addi- 
 tional Etruscan epitaphs, in particular, are of the 
 highest importance, and have induced me to return 
 once more to the subject of the Etruscan language; 
 for it does not appear, at least to my knowledge, that 
 all the results to which the new inscriptions lead have 
 hitherto been deduced from them. What the signifi- 
 cance of those results is may be estimated from the 
 facts, that we are now in possession of the written 
 forms of several Etruscan numerals, that we are enabled 
 to recognise the Etruscan equivalents for the Latin 
 -ginta and -genti, and that, in addition to the pre- 
 viously known ril, " year," we may likewise elicit 
 the Etruscan words for ce month" and " day." 
 
 The objects which I have chiefly had in view in the 
 following pages are : to demonstrate the Armenian 
 character of the Etruscan language in as complete and 
 brief a manner as possible, and to present such a 
 vocabulary of Etruscan words as may be sufficient for 
 the interpretation of the common forms of expression 
 on the monuments of Etruria. It would increase the 
 interest with which those relics of an ancient nation 
 
11 PKEFACE. 
 
 are regarded, if the meaning of the inscriptions which 
 they boar could be understood, and we were thus quali- 
 fied to know what the Etruscans engraved upon their 
 offerings to the gods, and on the tombs of their dead. 
 At the same time, these votive and sepulchral forms, 
 when combined with the Inscription of Cervetri, and 
 with such other points of evidence as may conveniently 
 be introduced in connexion with these three bodies of 
 proof, seem quite sufficient to disclose the nature of the 
 language of the Etruscans, and thus to determine who 
 that enigmatical people really were. 
 
 But the origin of the Etruscans is a question that 
 does not stand alone, and that cannot be treated 
 without touching on several others which relate to an- 
 cient and prehistoric times. To one of these in parti- 
 cular I have endeavoured to direct inquiry. As the 
 Armenians, the last representatives of the old Thracian 
 race at the present day, have been neglected or over- 
 looked in all investigations relating to the affinities 
 of the Etruscans ; so too a similar extension, in remote 
 ages, of the race of nations now confined to the Cau- 
 casian regions, is a probability that has not been 
 sufficiently allowed for in constructing the population 
 of Europe before the Aryans entered it. There seems, 
 as far as I can judge, to be no necessity for inferring 
 the extinction of either of the two ancient stocks 
 mentioned eighteen years ago by Dr. Latham in the 
 following passage, which sets forth very clearly and 
 forcibly the two principal questions whose solution 
 I have attempted : — 
 
 " The displacements effected by the different Euro- 
 
PREFACE. Ill 
 
 pean populations, one with another, have been enor- 
 mous. See how the Saxons overran England, the 
 Romans Spain and Gaul. How do we know that 
 some small stock was not annihilated here ? History, 
 it may be said, tells us the contrary. From history 
 we learn that all the ancient Spaniards were allied to 
 the ancestors of the Basques, all Gaul to those of the 
 Bretons, all England to those of the Welsh. Granted. 
 But what does history tell us about Bavaria, Styria, 
 the Valley of the Po, or ancient Thrace ? In all these 
 parts the present population is known to be recent, 
 and the older known next to not at all. The recon- 
 struction of the original populations of such areas as 
 these is one of the highest problems in ethnology. 
 To what did they belong, an existing stock more 
 widely extended than now, or a fresh stock altogether?" 
 
 " My own belief is, that the number of European 
 stocks for which there is an amount of evidence suffi- 
 cient to make their extinction a reasonable doctrine is 
 two — two and no more ; and even with these the doc- 
 trine of their extinction is only reasonable" 
 
 " a. The old Etruscans are the first of these"; 
 
 " b. The Pelasgi the second." 
 
 If the Etruscans were of the same race as the 
 Armenians, and the Pelasgi of the same race as the 
 Caucasians, both these stocks would still survive. The 
 Caucasian tribes include the Georgians (with the 
 Lazians and Mingrelians), who are connected through 
 the Suanians with the Circassians and Abkhasians; 
 the Ossetes, who are frequently supposed to be Aryans; 
 the Kisti, who are connected through the Tuschi and 
 
IV 
 
 PREFACE. 
 
 Pschawi with the Georgians ; and the Lesgi in the 
 ancient Albania. The languages of these tribes differ 
 very considerably, but something common is found to 
 run through them all. 
 
 The following equivalents will be employed for the 
 Armenian alphabet; and the Greek letters which corre- 
 spond to the Armenian in place, though not always in 
 sound, are prefixed to them : — 
 
 p. m. 
 
 a. a. 
 /3. b. 
 
 7- <7- 
 8. el 
 
 e. e (e or ye). 
 £. z (English z). 
 7], e. 
 
 e (e mute) . 
 6. th (Hebrew teth). 
 
 z (French j). 
 i. i. 
 
 I 
 
 Ich (guttural). 
 
 z {els). 
 k. k 
 
 It. 
 
 z {is). 
 X. t (Welsh 11: Polish t). 
 
 g {elz : English j). 
 
 v. n. 
 
 f . s (English sh) . 
 
 o. o. 
 
 c {ts : English ch) . 
 
 7T. p. 
 
 § {els: as sch in mensch). 
 p. r (strong r). 
 a. s. 
 
 iv (strong v). 
 t. t. 
 
 r. 
 
 z {tz : Hebrew tzacldi). 
 v. v. 
 
 (j>. ph {like p^h). 
 %. ch (Hebrew kopli). 
 co. 6 (broad o, or au). 
 
 f (used in foreign words). 
 
 Aemenian Diphthongs. 
 
 ov y vowel n> English oo. 
 
 ow, long o. 
 
 ea, like French e. 
 
 av } the older form of 6. 
 ev, like English yew. 
 iVj vowel y, French u. 
 
 The Armenian alphabet was invented about 1500 
 
 years ago. 
 
THE ASIATIC AFFINITIES OF THE 
 OLD ITALIANS. 
 
 CHAPTER I. 
 
 Migrations of the Aryans from their original home in Mount 
 Iniaus.— Eoute of the Thracians into Europe,— Western and 
 Northern limits of the Thracian area in Europe.— With what 
 nations the Thracian race came into collision.— Primitive popu- 
 lation of the South of Europe, and of Asia Minor and Armenia. 
 
 As the Asiatic element in Italy was mainly Etruscan, 
 and the Aryan character of the Etruscan language will 
 be apparent as soon as its examination is commenced, 
 such an examination may be appropriately preceded 
 by a sketch of the probable course of the Etruscans 
 from the original home of the Aryan family of nations, 
 and by an endeavour to determine what were the 
 elements which composed the early population of the 
 South of Europe. 
 
 A remarkable light has been thrown on the first 
 movements of the Aryans by the researches of German 
 scholars, the result of which is readily accessible in the 
 third volume of Bunsen's Egypt's Place in Universal 
 History, But the map which is there given as illus- 
 trating the " track of the Aryans from the Primeval 
 Country to India/' might perhaps receive with more 
 
 B 
 
2 THE ASIATIC AFFINITIES OF 
 
 justice the title of "a map to illustrate the tracks of 
 the Southern Aryans from the Primeval Country to 
 India and Armenia" The " Primeval Country" was the 
 mountainous region which contains the sources of the 
 Oxus and the Jaxartes : and the map exhibits clearly 
 how the Aryans, starting from this country, settled 
 successively in Sogdiana, Margiana, Bactria, Parthia, 
 Aria, and other tracts, until the list of the districts 
 which they occupied in the earliest times concludes at . 
 last with these three countries : — 
 
 14. Varena, now Ghilan, S.W. of the Caspian. 
 
 15. Hapta-Hindu, now the Punjab. 
 
 16. "The sixteenth country has no specific name. 
 Its inhabitants are the dwellers near the sea-coast, who 
 do not require any ramparts. Their curses are winter 
 and earthquakes. As the Caspian was the sea nearest 
 to the Old Iranians, we must understand the shores 
 of that sea." 
 
 It seems a highly probable inference that this last 
 country was Armenia, which formerly touched Ghilan 
 and the Caspian Sea* is protected by the natural 
 yparts of its mountains, has a long and severe 
 winter on account of its elevated position, and is 
 notoriously subject to earthquakes.^ But, even with- 
 
 * See the map in Whis ton's Moses Choronensis. 
 
 t " In the summer of 1840 Armenia was visited by a violent 
 earthquake, which shook Ararat to its foundation. The immense 
 quantities of loose stones, snow, ice, and mud, then precipitated 
 from the great chasm, immediately overwhelmed and destroyed 
 the monastery of St. James and the village of Arghuri, and spread 
 destruction for and wide in the plain of the Araxes. Although 
 
THE OLD ITALIANS. 3 
 
 out this inference, the mention of Ghilan immediately 
 before the Punjab, and the positions of the other 
 thirteen regions previously named, would lead up to 
 the following conclusion : — 
 
 The Southern Aryans, proceeding from the banks 
 of the Oxus, and expanding as they advanced, reached 
 Armenia on the west about the same time as they 
 occupied the Punjab on the east, and before they 
 entered Southern Media, Persia, Carmania, Gedrosia, 
 and India beyond the Sutlej. 
 
 Whatever may be the historical value of these 
 results thus deduced from the Vendidad, they fall in 
 at any rate singularly well with the theory which I 
 desire to support, and which may be stated in this 
 manner : — 
 
 The Southern Aryans were ultimately divided into 
 three principal stocks : the Thracian on the west, the 
 Medo-Persian in the centre, and the Indian on the 
 east. The Thracian race, as a distinct member of the 
 South Aryan family, had its origin in Armenia, about 
 the same time, and in the same manner, as the Indian 
 race had its origin in the Punjab. Finally, while the 
 Medo-Persians were gaining possession of the southern 
 half of Iran, upon the Indian Ocean and the Persian 
 
 Ararat is formed of volcanic rocks, yet no allusion to its volcanic 
 activity at any period, no mention of an eruption, is made by any 
 of the native historians, who record, nevertheless, several earth- 
 quakes more or less calamitous." Appendix to Cooley's transla- 
 tion of Parrot's Journey to Ararat, p. 371. The earthquake of 
 1840 was felt as far as Tiflis, 150 miles N. of Ararat, and as Tauris 
 or Tabreez, 150 miles S.E. of Ararat, near Ghilan. 
 
I TIIE ASIATIC AFFINITIES OF 
 
 Gulf, and while the Indians were carrying the Sanskrit 
 language with them from the Indus to the. Bay of 
 Bengal, the Thracians were extending themselves from 
 the Caspian to the Alps and the Tyrrhenian Sea, and 
 carrying an Armenian dialect into Etruria and Rhcetia. 
 That there were Thracians over all this extent of 
 country is, however, not merely a probable or possible 
 conjecture: it is a matter of ancient history, or at least 
 of ancient tradition. What the earliest Zoroastrian 
 record seems to exhibit in the germ, the authors of 
 Greece and Koine present in its completion. Twenty- 
 two such authors, as Mr. Dennis has noticed, derive the 
 Etruscans from the Lydians; a presumption of affinity 
 not to be hastily set aside, although the voyage of the 
 Lydians to Btruria under the conduct of Tyrrhenus 
 may be no more historical than the voyage of ^Eneas, 
 and Tyrrhenus himself a personage like Hellen and 
 Romulus, or Delphinus and Sabaudus, the sons of 
 Allobrox. That some Alpine nations, and especially 
 the Rhaotians, were akin to the Etruscans, is a fact 
 attested by Livy : and that the Lydians and Carians 
 were allied in blood and language to the Mysians, who 
 were a branch of the Thracian race, is affirmed by 
 other writers. There were too, as we learn from 
 Strabo, Thracians mixed with the Celtic inhabitants of 
 Noricum and Pannonia, the countries which intervene 
 between Rhaotia and Dacia. The rest of the historical 
 inicnt for the extension of the Thracians from 
 Armenia to Italy may be summed up in the words of 
 Dr. Latham, although he has rejected the result which 
 
THE OLD ITALIANS. O 
 
 I not only accept, but extend to Etruria :* " The old 
 Thracian affinities are difficult, but not beyond in- 
 vestigation. A series of statements on the part of 
 good classical authors tell us, that the Daci were what 
 the Getae were, and the Thracians what the Geta3; 
 also, that the Phrygians spoke the same language as 
 the Thracians, and the Armenians as the Phrygians. 
 If so, either the ancient language of Hungary must 
 have been spoken as far as the Caspian, or the ancient 
 Armenian as far as the Theiss" Write here the 
 Alpine Rhine and the Tiber for the Theiss, and I believe 
 that no more than the truth would be said, and perhaps 
 not quite as much as the whole truth. For I imagine 
 that the Bebryces, whom several authors mention in 
 the Eastern Pyrenees, were Thracian settlers who 
 came thither by sea, probably from Italy, before the 
 Carthaginians and Greeks formed settlements upon 
 that line of coast. My reasons for this conjecture may 
 be thus briefly expressed, as a part of the cumulative 
 proof of the western extension of the Thracian race : — 
 sarn, ""ice:" root sar, "freeze." 
 patel, " to enclose" (th. pat) . 
 L patovar, " wall, rampart." 
 
 Bithynia ... Patavium, a town of the Thracian 
 
 Bebryces. 
 Pannonia ...... Patavium, a town, now Pettau. 
 
 Armenian . . . - 
 
 * Ethnology of Europe, p. 229 (1852). Dr. Latham considers 
 that two languages were spoken in Phrygia ; one allied to that of 
 the Armenians, and the other to that of the Thracians, whom he 
 regards as Slavonic. 
 
6 THE ASIATIC AFFINITIES OF 
 
 Venetia Pataviumythe chief town, now Padua.* 
 
 E. Pyrenees .. .Bebryces. -\ 
 
 Pyrenees sern-eille, " glacier." For the termi- 
 nation, compare ab-eille, sor-ella, 
 and or Mia. The rest is Armenian. 
 Observe, too, that Iberians bordered 
 on these Bebryces, and on the Ar- 
 menians. 
 Patavium, the chief t ; own of Venetia, seems to have 
 a Thracian and Armenian name. Of the Venetian 
 language I know only one word, which is given by 
 Pliny (H. N., xxvi, 6) : "Halus antem, quam Gralli sic 
 vocant, Veneti cotoneam." Ootonea " comfrey, gv/jl- 
 (f>vrov, wallwurz," may be explained, like the Dacian 
 Koriara, " aypooaTts, gramen," from the Armenian 
 hhotj " herb, forage," an Armenian word which nearly 
 replaces the English wort and the German wurz in 
 names of herbs. The Armenian Tchotan, "low, humilis," 
 shews a connexion in sense between hhot and humus. 
 Ptolemy mentions an Armenian town called Koracva. 
 The Cot-ensii were a Dacian tribe. 
 
 The Bebryces in Roussillon, with the word sern- 
 eille, " glacier," would mark the extreme western ex- 
 tension of the Thracians. On the north-west their 
 limit would have been Kh83tia, where their presence is 
 indicated by several Rhaeto-Bomance words used in 
 the Swiss Canton of the Grisons. The following group 
 
 * These three Patavia, and no others, are mentioned by Ptolemy, 
 f See Bouquet, Historiens de la France, vol. i, pp. 94, 114, 531, 
 677. 
 
THE OLD ITALIANS. 
 
 of five kindred terms may mark how the Southern 
 Aryans once reached from the Ganges to the sources 
 of the Rhine, while the existence in Lydian of the 
 termination of the Armenian present participle, -avt, 
 6t, or ot, is one sign that the Lydians had a more 
 intimate degree of affinity with the Armenians than 
 with the Indians. The Etruscan language, when we 
 come to examine it, will exhibit exactly the same 
 degrees of affinity to the Armenian and the Sanskrit 
 that the ancient Lydian and Rhgetian appear to have 
 possessed. These are the five words : — 
 
 Sanskrit kantha, " the throat, the throttle." 
 
 Armenian khetdavt, " throttling, choking." 
 
 Lydian fcav8av\-r]<;, " o-fcvXko7rvi/CT7]<;, the 
 
 quinsy/ 5 * 
 
 Albanian kyendis, " I choke. "f 
 
 Rhseto-Romance. . candarials, ~ a choking disease." J 
 
 The following Rhaeto- Romance names of animals 
 exhibit also Armenian affinities : — 
 
 Guis, " marten." Armenian Ttovz, " pole-cat ;" 
 hznachis, "marten," = Polish and Bohemian kuna, 
 Russian kuniza, Lithuanian kiaune. 
 
 Asoly asoula, "kid." Armenian ayz, "goat" (= Sans- 
 krit affd, Greek .cut;) ; ovl, "kid." 
 
 Tama, "moth." ) Armenian tliithern, 
 
 Fafarinna, "butterfly." J "butterfly." 
 
 # Bottioher's Arica, p. 44. f Hahn's Albanesische Studien. 
 
 X "Eine Art Driisenubel, das das Athnien sehr erschwert." 
 Carisch^s Rhato-Romanisches Worterbuch. 
 
THE ASIATIC AFFINITIES OF 
 
 Salipp, "locust." Sanskrit galabha, "locust:" 
 root ra 1, " to run." Armenian satap, " quick, gliding." 
 
 There are thus signs of the Armenian language 
 having once stretched as far as the Pyrenees and the 
 Alps : and the same may be said of the Carpathians, 
 for the relics of the Dacian language exhibit some 
 striking instances of Armenian affinity. These relics 
 consist of more than thirty names of plants used in 
 medicine ;* names that are very likely to contain the 
 Dacian equivalents for the German kraut and wurz, and 
 the English grass, wort, or weed, which are the terms 
 that most commonly enter into the composition of 
 German and English names of plants. The correspond- 
 ing terms in Armenian are : Jchot, " herb, verdure, 
 hay," and det, " herb, medicine, poison/" Thus 
 "tobacco" is zJchalchot, "smoke-herb," and "rhubarb" 
 is lihasndet, "flock-wort," in Armenian. Are there 
 any indications of kindred words in Dacian ? 
 
 Now aypcQcrri*;, gramen, was called in Dacian KOT-iara 
 or /coT-TJara, in which we may fairly recognise the 
 Armenian hhot, "herb, hay;" while it is very probable 
 that a word similar to the Armenian det, "herb, 
 medicine, poison," existed in the Dacian SteX-eta or 
 SteX-Xeiva, "henbane;" rev-ScXd or rev-hetXd, " cala- 
 lnint;" hovco-hrjka, " origan ;" irpia-hrjXa or irpia- S iXd, 
 " black briony ;" Koiico-hCkd (or possibly tcvfco-XlSa), 
 " nightshade;" and perhaps irpoire-hovXd or irpoire-htXd 
 " cinquefbil." In addition to Idiot, the Armenian has 
 
 * Grimm, Geschichte der Deutschen Sprache, cc. 9, 30. 
 
THE OLD ITALIANS. 9 
 
 another word for " grass/' sez, apparently = Sanskrit 
 qaka, "herba." This word, combined perhaps with 
 the Armenian anyag, " unlucky, detrimental/' may be 
 found in the Dacian aviaaae^e, " onobrychis :" Pliny 
 (xxiv, 113) speaks of a plant called impia herb a. 
 There is too in Armenian the word ost, "ramus, 
 germen, palmites, frondes" (cf. Basque ost, "leaf/' 
 and German ast), which, when combined with the 
 Armenian zow, " QaKaaaa^ gives a good explanation 
 of the Dacian ^ovoarrj, "artemisia," of which Dioscorides 
 says that it grows for the most part ev 7rapa6aXa<TaloL<$ 
 toitols. There is a fifth Armenian word of the same 
 class as Mot, det, sez, and ost, which are all apparently 
 found in Dacian. This word is phthith, " the blowing 
 of a flower," which gives the verb phtheth-il, "to 
 blow, to bud, to sprout," with the present participle 
 \phtheth-ot, the preterite participle phtheth-eal, and the 
 future participle phtheth-eli. We meet likewise with 
 phthith, when combined with zatih, "flower/' mist, 
 ;" always," and loys, "light," in the following com- 
 pounds : — 
 
 zatk-a-phthith, " flowering, blooming." 
 mest-a-phthith, " ever-blooming." 
 lovs-a-phthith, " luminous, light-shedding." 
 To these add one of the previous Armenian I termina- 
 tions, or such a one as in ovs-et, "sensible," from 
 \ovs, " sense," and then compare the Dacian— 
 <\>i6-o-$6e6eka, " ahiavrov, maidenhair."^ 
 
 <M- is perhaps to be found in the Armenian %>het-ovr, "feather/* 
 
10 THE ASTATIC AFFINITIES OF 
 
 It lias been pointed out by Grimm that one Dacian 
 word, Kpovardvrj, u ^e\iS6vtov /Jbeya/' is the Lithuanian 
 hregzdyne, and that another, hvv, "urtica, KvlSr)," is 
 the Welsh dan-ad. The existence of such words in 
 Dacian may, however, be accounted for by contiguity 
 of position; an expedient which cannot be admitted as 
 an explanation of the Armenian affinities of the Dacian, 
 which are, besides, more numerous and intimate than 
 any other. There were Celts in Pannonia, and there 
 may have been Lithuanians in Galicia ; but Armenia 
 is far away from Hungary and Wallachia. 
 
 Having now ascertained, by the combined aid of 
 history and language, the probable limits of the Thracian 
 area in Europe, we must proceed to consider another 
 subject before entering upon the examination of the 
 Etruscan language. What nations possessed the area 
 in question when the Thracians first intruded upon it? 
 
 I spoke of Bunsen's map {ante, p. 2) as illustrating 
 the tracks of the Southern Aryans to India and Armenia; 
 for it is hardly probable that all the Aryans entered 
 Europe through Armenia and Asia Minor. If the 
 €€ Primeval Country" of the Aryans was the region 
 where the Oxus and the Jaxartes have their sources, 
 then another branch of that race, who may be called 
 the Northern Aryans, would most likely take their 
 way into the West along the north of the great 
 barrier of sea and mountain, a thousand miles in 
 
 and phet-tel, "to pluck." $>ieo<p6e0e\d would then be "feather- 
 sprouting, plumy," just as the Armenian lovsaphthith is "light- 
 sprouting, luminous." 
 
THE OLD ITALIANS. 11 
 
 lengthy which, is formed by the Caspian, the Caucasus, 
 and the Euxine. The leaders of this division seem to 
 have been the nations of the Classical stock, such as 
 the Umbrians, the Oscans, and the Hellenes ; which 
 last I think, with Dr. Latham, to have been maritime 
 settlers from Italy, as the Pelasgians and the Leleges 
 may have come from Asia. I should find the explana- 
 tion of Pelasgi, who are described as veteres, as 
 avTO'xOoves , as ap^aiorarot, and as apyulov tl fyvkov, in 
 the Armenian words, wat azg, " apyalov cf>v\ov;" where 
 azg is "race, nation/' and wat is "ancient, old," 
 = Greek TraX-aios, = Epirot ireX-ios. Pelasgi would be 
 a Thracian term corresponding to the Greek Auto- 
 chthones and the Latin Aborigines. Armenian words 
 similar to Pelasgus in formation are: lavazgi, "noble," 
 = good-race ; watazgi, " plebeian," = bad-race ; 
 azatazgi, "citizen," = free-race; aylazgi, "foreigner," 
 = other-race ; and some more. Leleges, " again," is 
 readily explained as " sailors," from the Armenian let, 
 lot, lovt, "swim," lotah or lovt ah, "swimmer;" nava- 
 lovtak, "navigating."* The Classical nations would 
 probably, as might be expected from the level nature 
 
 * " The headlands of Southern Greece, and some other parts of 
 the coast, were occupied in the earliest times by the Leleges and 
 other tribes, which spread themselves from the opposite shores of 
 Asia Minor over the islands of the Moean Sea. But, with these 
 exceptions, the whole continent, from the borders of Thrace and 
 Macedonia to the extreme point of Peloponnesus, was peopled by 
 the great Pelasgian nation" (Maiden). If the Leleges came by 
 sea into Greece, and the Bebryces by sea into Eoussillon, the 
 Tyrrhenians might have come by sea into Etruria. 
 
1 2 THE ASIATIC AFFINITIES OF 
 
 of their supposed route from the Oxus to the Danube, 
 reach Italy before the Thracians, while the reverse 
 may have been the case in Greece; and they would 
 have been followed or accompanied into Europe by 
 their kindred the Celts, till the Alps, or perhaps the 
 Carpathians, severed the stream into a northern and a 
 southern arm. The Celts, in their turn, would have 
 been followed by the Germans, and the Germans 
 finally by the Slavonians and Lithuanians; although it 
 is possible that these two ramifications of the Sarma- 
 tian branch of the Aryan stock, whose languages have 
 several Armenian affinities, may have preceded the 
 Thracians through Asia Minor. I do not, however, 
 think so myself. If the Thracians entered Europe 
 from Asia Minor, and the Celts through the South of 
 Eussia, it might be anticipated that the two races 
 would clash and mingle on the Lower Danube; and 
 this would account, not only for what Celtic may 
 appear in Dacian, but also for such Celtic words as the 
 Etruscan has taken up. 
 
 But other tongues, besides the Celtic or any other 
 Aryan language, may have affected the original Thracian 
 as it was carried from Armenia to Etruria. Europe 
 would have been peopled by some other nations before 
 the Aryans entered it. Now, if we eliminate from 
 Europe, together with that part of Asia which lies 
 between the Caspian and iEgean Seas, all Aryan, 
 Semitic, and Turkish inhabitants, we shall be left 
 with only three races, or groups of nations : the 
 Basques in the west, the Fins in the north, and the 
 
THE OLD ITALIANS. 13 
 
 nations of the Caucasus in the east. The first ap- 
 proximation therefore that we should make towards 
 reconstructing the primitive population of Europe 
 (with Armenia and Asia Minor) would be, by extend- 
 ing the Basques, the Fins, and the Caucasians, till 
 they met somewhere in the centre of Europe. I do 
 not, however, find any linguistic signs of a Basque 
 population there, though the Taurine asia, " rye," has 
 been compared with the Basque acia, ""seed;" and the 
 name of the Taurine Iria, now Voghera, is like the,/ 
 Basque iria or utia, "city." But I must notice \ 
 three Alpine words which deserve attention as possible 
 indications of an extension of Fins and Caucasians 
 into those mountains. The Ossetes and Tuschi, it has 
 been mentioned in the preface, are two Caucasian 
 tribes; and the Tuschi would apparently be a remnant 
 of those Tusci whom Ptolemy speaks of in Asiatic 
 Sarmatia on the north slope of the Caucasus. It 
 might be an extension of the same race that brought 
 the Tuscan name into Europe, where we meet with it 
 in Etruria. The three Alpine words are : — 
 
 1. Kdss, fcees, lease, "glacier" (Noric Alps).* Lap- 
 ponic Jcaisse, "mons altior, plerumque nive tectus." 
 Esthonian kahho, "frost; hassejda (jaa, "ice"), "ice 
 formed by frost upon snow." Georgian qisiva, " frost. "f 
 
 2. Laid, lauwi, lawine, "avalanche. "J Tuschi law, 
 " snow" (Schiefner's Thusch-Sprache). 
 
 * The Noric Alps lie to the north of the great valley of the 
 Drave, the Carnic Alps to the south. 
 
 f Compare the Peruvian cassa, "hail/ 5 Helps' Life of Pizarro. 
 J In Bhseto-Romance, lavinna. 
 
14 THE ASIATIC AFFINITIES OF 
 
 8. Tiissel, "boy, knabe" (Spliigen district). Carian 
 rovcravXoL, " Hvy/JLaloL." Ossetic tyiisul, " little/' 
 = Armenian doyzn. Armenian thzovk, " a pygmy/' 
 root thiz, "a span/' Kdr-rov^a, 7t6\l<; <dpatcr)<; } ev § 
 tcarcpKovv oi Uwy/jLaloL* 
 
 The last of these three words may possibly be 
 Aryan, and have been carried into Rhastia by the 
 Thracians; but of the other two, one seems Caucasian, 
 and the other both Finnish and Caucasian. There 
 would be two explanations of the presence in the Alps 
 of Turanian words for "ice" and " snow/' if it may be 
 allowable to apply the title " Turanian" to the Fins 
 and Caucasians. Such words may have been taken up 
 in Asia by the Thracians, and have been brought by 
 them to Rhaetia ; or they may be due to an original 
 Turanian population extending from the Ural or the 
 Caucasus to the Alps. If we adopt, as I should be 
 inclined to do, the second of these alternatives, we 
 might be led to inquire if there were any European 
 nations in historic times who belonged to this primi- 
 tive Turanian population. That the Basques or Iberians 
 formed a part of it is no improbable supposition ; and 
 I suspect that both they, and all the original inhabit- 
 ants of the South of Europe, as well as of Asia Minor 
 and Armenia, were allied in blood and language to the 
 Caucasian nations, while the North of Europe, beyond 
 the Alps and the Carpathians, would have been Fin- 
 nish, f The identity of names is remarkable. Not 
 
 * Botticher's Arica, p. 5. 
 
 t " The population of the first period," says Mr. Troy on, in his 
 
THE OLD ITALIANS. 15 
 
 merely were there Iberians and Tuscans in the Caucasus 
 as well as in Spain and Italy ; but the name of the 
 nation which intervened in Europe between the Iberians 
 and Tuscans, namely, the Ligyes or Ligurians, is found 
 likewise in or near the Caucasus, as we know from 
 Herodotus and Zonaras. There were, too, Ligyrii in 
 Thrace, and the Taurisci of Noricum bore the additional 
 name of Ligyrisci. The Ligurians of Italy were almost 
 extirpated in their long war with the Romans (Niebuhr), 
 and both in Italy and Gaul they would have been 
 much Celticised, as we know that the Salyes were; 
 while the Tuscans of Italy, if originally Caucasian, 
 would have been conquered by the Thracian Rasenee, 
 and have lost their national existence, while they 
 communicated, like the Britons, their name to their 
 conquerors. The same lot probably befell the primi- 
 tive population of Dacia and Mgesia ; for, though the 
 Dacian language was Thracian in the time of Diosco- 
 rides, as may be inferred from its Armenian affinities, 
 and as Strabo had declared at a rather earlier period, 
 yet the Dacian town-name dava, of which a few 
 
 j work on the Swiss Lake-dwellers, f< are a primitive people, per- 
 1 haps belonging to a Finnish or Iberian race which came out of 
 
 Asia several thousand (hundred) years before our era, and folio w- 
 : ing the course of the Ehone or the Rhine wandered into the 
 
 valleys of the Alps." Keller's Lake Dwellings (Eng. trans.), p. 395. 
 ! I imagine that law-ine and Jcees are relics of the language of this 
 j Finnish or Iberian race, using Iberian in the double sense of 
 
 Georgian (or Caucasian) and Basque; like as the ancient presence 
 
 of the Thracian Bebryces in Eoussillon would be indicated by the 
 
 word sern-eille, " glacier," in the Pyrenees. 
 
16 THE ASIATIC AFFINITIES OF 
 
 instances are found in Msssia besides, with one in 
 Dalmatia, and one in Dardania, seems best explained 
 from the Georgian data, "village, town/* though 
 probably allied at the same time to the Sanskrit dega 
 and the Armenian deh, " pagus." A parallel might be 
 made with Cornwall, where Celtic place-names exist 
 among a people now speaking English. The Dad or 
 Davi were probably " the villagers/' as Gesenius has 
 interpreted the name of the Caspian Dahce or Dai.* 
 
 # The Georgian daba might also explain the name of the Gala- 
 tian Tavium, and the last part of Pa(t)-tavium (ante, p. 5), while 
 Pat- was referred to the Armenian pat-el, "to enclose/' There was 
 a Dacian town called Patav-issa. Two more Dacian towns were 
 called Acidava,, which bears a resemblance to Acitavones, the 
 second name of the Centrones in Savoy. We know that there 
 were the pile-towns of an early race on the Leman Lake — 
 
 ' ' Quam vetus mos Greecise 
 Yocitavib Action." (Avienus) 
 
 Guadix in Andalusia was anciently called Acci. In Georgian, vake 
 is "plain, fields," like the Spanish vega. In Lapponic, wagge is 
 " valley :" " imprimis vallis inter montes latior," which would well 
 describe the position of the Ligurian Vagienni. The Genoese 
 lalla, "aunt," seems Finnish too, as lei is "uncle" in Esthonian. We 
 find in Avienus the inhabitants of the Alpine valley of the 
 Ehone called Tylangii, Daliterni, Chabilci, and Temenici, instead of 
 Viberi, Seduni, Veragri, and Nantuates ; just as the Acitavones were 
 also Centrones, and as the Medulli were also Go.ro celi. Each one of 
 the six tribes between the Furca and the Mont Cenis had two 
 names, and so likewise had the Lacus Lemanus. The Alps were 
 perhaps not much Celticised before the movement which brought 
 the Gauls into Italy. The Chabilci seem to have left their name 
 to Chable and the Chablais, which once included the Lower Vallais. 
 There was a town called Chabala in the Caucasian Albania 
 Temnus was a mountain in Mysia. 
 
THE OLD ITALIANS. 1 7 
 
 A part of the Albanians, the probable representa- 
 tives of the ancient Illyrians, are still called Toscans ; 
 and both are Caucasian names, though the Albanian 
 language is Aryan, as Bopp has shewn in his essay on 
 that subject. Among many remarkable Albanian words, 
 two may be selected on account of their significance. 
 One of them, diel or dil, "sun," resembles the Georgian 
 dili or dila, " morning/' but would also be allied in 
 root to the Albanian di(te), "day," = Latin dies, 
 == Sanskrit di(na), = Welsh dyiv, = Armenian tir. 
 The Armenian plural form, ticli, " age," i. e. " days," 
 would imply ti in the singular as another form of ti%\ 
 The Sanskrit diva," heaven," is likewise a kindred term; 
 and the root is the Sanskrit div, "to shine/' By 
 taking this root into the Armenian, dropping the v as 
 in the Sanskrit dina = div ana, and adding one of the 
 Armenian I terminations {ante, p. 9), the Albanian 
 diel or dil, " sun," comes to mean " shining, bright." 
 Yet I terminations are as characteristic of Georgian as 
 they are of Armenian and Etruscan ; and dili or dila, 
 " morning," is a true Georgian derivative from the 
 root, di(v), "to shine." In Tuschi, too, " God" is 
 Dal, which might be deduced from a similar root. 
 
 For from the root div the Sanskrit derives also 
 deva, " God," = Armenian dev, " demon ;" which 
 words, when combined with the Armenian iverin, 
 "high" (shortened into iuern in ivem-akan, "celestial"), 
 lead us at once to the Albanian perndi or perendi, 
 " God." The root of werin is the Armenian wer, " on 
 high" (= Sanskrit para, "altus"), which gives in the 
 
 c 
 
18 THE ASTATIC AFFINITIES OF 
 
 comparative weragoyn, " superior/' and enables us to 
 explain the Phrygian fiepetcvvSat, " haljJLove*;" i.e. "the 
 supremo spirits;" or else as "the most high/' if h is 
 not radical. Berccyntlins was also a mountain.* 
 
 The Armenian for "God" is astovaz, — Zend agtvat, 
 { s existentia (a$tu) pra3ditus."t 
 
 The Etruscan words for "Deity" are Aryan, like 
 the Armenian and Albanian. They are, according to 
 Classic reporters, cesar, " deus," and alcroi, " QeoL" 
 Acs- or ak- is the Old Norse as, "deus," the Armenian 
 ays, " spirit, demon/' and probably also the Gaelic 
 aos, "fire, the sun/' from which by the addition of 
 fear, "bonus, vir," = Sanskrit vara, Armenian ayr, are 
 derived the Gaelic Aosar, and the Irish Aosar, Aesar, 
 "God," which Bopp compares with the Sanskrit igvara, 
 " dominus/' and ig, " dominari." The connexion be- 
 tween " dominus" and " sol" is shewn in the Sanskrit 
 ind, "dominus," = Irish ion, "sol." Yet we should 
 expect the Sanskrit ig to become ic (= ik) in Gaelic. 
 The Etruscan for " heaven," falandum, would be 
 allied to the Persian buland, "high, heaven," and the 
 Sanskrit oland, "in altum tollere." 
 
 Whatever the population may have been originally 
 in Illyria, Dacia, and Msesia, yet it seems eventually 
 to have been either Aryan or Aryanised, as it would 
 certainly have been in Thrace and Greece : but among 
 the mountains in the heart of the peninsula a remnant 
 of the original Turanian inhabitants may have been 
 
 * Other languages present signg.of affinity here. See Diefen- 
 bachj Lex, Comp. s.v.fairguni, "berg." 
 f Botticher's Arica, p. 63. 
 
THE OLD ITALIANS. 19 
 
 left in the Pseonians, with the dwellers on piles whom 
 Herodotus describes on Lake Prasias ; the kindred, 
 perhaps, of the Swiss lake-dwellers who have of late 
 excited general interest, and whom "-Realmah" has 
 made so popularly known. The Pseonians represented 
 themselves, according to Herodotus, as a colony of the 
 Teucrians from Troy, and their remains were certainly 
 called Gergithes (v. 22), and probably dwelt at Gergis, 
 Gergithium, or Gergithus, in the territory of Lampsacus 
 (Strabo, p. 589). Gergeti is an Ossetic town in the 
 centre of the Caucasus. The pile-dwellers on Lake 
 Prasias cut their timber in Mount Orbelus, and the 
 Orbelians were a princely family in Georgia.* 
 
 If there were people of Caucasian origin in Europe, 
 it is not likely that Asia Minor would be without 
 them. The most eminent nation here that I should be 
 inclined to consider as Turanian and Caucasian would 
 be the Lycians, whose language, which is neither 
 Aryan nor Semitic in character, appears to me to 
 present signs of such an affinity. I have discussed the 
 question at some length in my Armenian Origin of the 
 'Etruscans, and shall only select a very few points for 
 notice here. The Lycian term for " wife," lade, de- 
 serves consideration in the first place, and may be 
 thus derived from the Caucasian by the aid of the 
 Circassian and Lesgi languages : — 
 
 fe}" flesk " 
 
 Circassian ...-{lay l„ blood » 
 t lilay J 
 Jklay, " husband." 
 * St. Martin, Memoir es sur V Armenie. 
 
20 TTTE ASIATIC AFFINITIES OF 
 
 Lesgi tly'adi ) « wife » 
 
 Lycian lade ) 
 
 Three other Lycian words are : — 
 
 tede-eme, " son." 
 
 KciSp-e/Jia, " (jltov (ftpvy/JLOs" 
 
 vX-a/^os, " fcapTros." 
 
 the last pair being two of the five Lycian words 
 
 derived from Greek reporters. The terminations of 
 
 these three words may be found in the termination of 
 
 the Esthonian infinitive -ma, which resembles the 
 
 Turkish infinitive terminations, -mak, -mek ; or in the 
 
 Tuschi noun-termination -om, which is, however, not 
 
 common ; or in the Lapponic noun-termination -em, 
 
 which is exceedingly frequent : as, e. g. — passat-et, 
 
 "lavare/' passat-em, "lotio;" — passot-et, " colere/' 
 
 yassot-em, " cultus;" pass-et," ^ssaxe" pass-em , u Yevxi." 
 
 But Aryan languages have similar terminations, as in 
 
 av-ejjios. 
 
 For vX-cl/jlo? we have — 
 
 Turkish el-ma 
 Hungarian al-ma 
 
 and for v\- — 
 
 Esthonian willi 
 
 > " apple 
 
 Tuschi chit f 
 
 r . / khili, "apple." 
 Georgian j ^'.^ PP^ 
 
 for KciSp- — 
 
 Tuschi Icotor, " cake, bread." 
 
 and iur tede- — 
 
 Esthonian told- } ,, . , ,, 
 
 Armenian taz- / nounsh - 
 
 Greek r?]0r), " nurse/ 
 
THE OLD ITALIANS. 21 
 
 I subjoin one Lycian epitaph to shew the language, 
 
 and because it will enable me to explain a word 
 
 better than I did on a former occasion : — 
 
 ewuinu goru mute prinafatu esedeplume urppe 
 this tomb here made Esedeplume for 
 
 lade euwe se tedesaeme eaweye womeleye. 
 wife his and children his all. 
 
 For the last word compare : — 
 
 /wune a some, was, welches" 
 
 m i • J wun — ele ..." every, jedes." 
 
 \ wum " something, etivas." 
 
 vwuma "all, alles, alle" 
 
 Lycian . . .worn — ele — ye, " all." 
 
 On a gold collar about six inches in diameter, 
 
 which is now in the Museum of Vienna, and was 
 
 found in Wallachia, a part of ancient Dacia, in the 
 
 year 1838 (Micali, Mon. Ined., tav. liii), there is an 
 
 inscription written from left to right, which may be 
 
 read by the aid of the Lycian characters, alv.fi/hthdi 
 
 iifipfa, "the iifipfa of Alufiithas," a name partially 
 
 like Alyattes * Here the short a is X, which is 
 
 Lycian ; and the u, I surmounted by X, which is also 
 
 'Lycian. If the X be made a ^ or ch instead of a short 
 
 a, the inscription would become chlafiithdi iifipfch, 
 
 , which seems a less likely reading. The remaining 
 
 letters in the inscription have nothing particularly 
 
 distinctive about them. If the collar were a votive 
 
 offering, or even a gift to an individual, iifipfa might 
 
 * Pausanias (x, 16) says, that of all the offerings of the Lydian 
 kings nothing remained at Delphi but the iron pedestal or base 
 )f the bowl of Alyattes. 
 
22 THE ASIATIC AFFINITIES OF 
 
 be explained from the Tuschi vphob == Georgian iephoba, 
 ' € generosity, munificence, freigebigkeit ;" wliicli con- 
 sists of the Tuschi iaph, = Georgian iephi, " munifi- 
 cent/' combined with the Tuschi termination -ob 
 = Georgian -oba. Votive offerings are said to be 
 given Jibe utcr and libera munere. Such a Georgian 
 form as saso-cba, "hope/' comes still nearer to tif* 
 ipfa, though -ipfa should likewise be compared with 
 the Sanskrit gal-abha, = Armenian sat-ap, — Rhaeto- 
 Eomance sal-ipp (ante; p. 8).* We have seen from 
 the Dacian -clava, = Georgian daba, "village, town/' 
 that the language of Dacia was probably at first 
 Caucasian, and from the Dacian names of plants that 
 it became afterwards Thracian ; as the language of 
 Gaul became Latin while the Celtic town-names were 
 retained in the country. Yet it does not follow, 
 because the gold collar was found in Dacia, either 
 that its original destination was Dacia, or that the 
 inscription on it is Dacian. It might have been carried 
 off, and brought to Dacia. For the declension of 
 Alufiithdi, compare the Tuschi Marital, the genitive of 
 Marie, "Marcus/' and masai, the genitive of rnasa, 
 " lux/' = Sanskrit mds, " luna," = Georgian mze, 
 "sol." 
 
 Micali thinks the characters on the Wallachiau collar 
 to be like the Euganean letters. The Euganeans may 
 very well have been a remnant of the early Turanians 
 of Italy. They were neither Gauls nor Venetians. 
 
 Compare apfel, hihpfen, kopf, and apple, hop, cop. 
 
 
THE OLD ITALIANS. 23 
 
 Maiden writes, in his unfinished, and hardly com- 
 menced, History of Rome: "The mountainous country 
 northward from the lake (of Grarda) remained in pos- 
 session of the Euganei. Of this ancient and once 
 powerful people Cato was still able to enumerate thirty- 
 four towns (Plin. H. N. in, 24) ; and they were re- 
 ported by tradition to have inhabited all the country 
 between the Alps and the Adriatic Sea, till they were 
 driven into the mountains by the Yeneti (Liv. i, 1). 
 Their chief tribe was the Stoni or Stoeni (Plin. R. N. 
 iii, 24), and the Stoeni are expressly named Ligurians 
 in a fragment of the Triumphal Fasti, and by the 
 geographer Stephanus." I think it probable that the 
 Euganeans are rightly connected with the Ligurians, 
 as well as the Orobii, whose origin Cato could not 
 ascertain (Plin. H. N. iii, 23). The same mystery 
 hangs over the origin of the Euganeans. Micali says 
 (vol. ii, p. 25): " Vanamente pero vorremmo rintracciare 
 Torigine degli Euganei." No doubt it is difficult to do 
 so. Indeed, until Caucasians or Pins be brought in, 
 there is always one race at least, in Italy, in the 
 Turkish peninsula, and in Asia Minor, which cannot 
 
 , be accounted for. The name of the Orobii, who occupied 
 the mountains of Como, and possessed Berg-omum, 
 (which is like Perg-amus, and may contain a termina- 
 
 ; tion like the Tuschi -om and that of the Lycian iJX-a/^o?) ; 
 
 i of the river Or obis, now the Orb in Languedoc ; of the 
 river Orba in the Ligurian Apennines; of the Pgeonian 
 Mount Orbelus ; and of the Georgian Orbelian family; 
 —all these names may be compared with three given 
 
2 I- THE ASIATIC AFFINITIES OF 
 
 by Ptolemy: Orbanassa in Pisidia, and Orbisene and 
 Orbalissene in Armenia Minor, where -sens would be 
 the Armenian termination sen, found in mezasen, 
 "great," and sikasen, "red." Cf. Tyr-senus and Por- 
 . It signifies also "town," and is Georgian as 
 well as Armenian. Orbi means " eagle" in Georgian, 
 and ~((H and -eZt, which may appear in Orb-elus and 
 Orb-aKssene, are common Georgian terminations: e. g. 
 dab-all, "humilis," from dab-a, "pagus;" Guri-eli, 
 " belonging to the country of Guria." Mount Orbelus 
 would probably = Adlersberg, but the rivers Orba and 
 rob is would rather mean "swift," like as the Etruscan 
 aracus, "hawk/' is the Armenian arag, "swift." Cf. 
 Araxes. There is an Arbel-lwrn in the Bernese Alps. 
 
 The Tuschi call their own country Baza (Schiefner, 
 s. v.) ; a name resembling the Hispanian Basti, now 
 Baza, which Humboldt compares with the Basque 
 laso(a), "forest." The Bessi were a Thracian tribe. 
 
 There was an ancient people in Paphlagonia or 
 Bithynia called Caucones, who were extinct in Strabo's 
 time. Some thought them to be Scythians, others 
 Macedonians, and others Pelasgians. A people of the 
 some name were once found in Messenia and Blis. 
 The Gauco-ensii were a Dacian tribe. Cf. Caucasus. 
 
 In addition to the Lycians, I should be inclined to 
 consider as Caucasian the Pisidians and Lycaonians, 
 neither of whom are mentioned as Thracians, though 
 the Phrygians and Milyans, who bordered on them, 
 ore expressly said to be so. Lycaon was a son of 
 P lasgus. When the Thracians advanced from the 
 
THE OLD ITALIAN'S. 25 
 
 Caspian into Armenia and Asia Minor, I think that 
 they left the relics of the primitive Caucasian popula- 
 tion, on their right in Armenia, and on their left in 
 Asia Minor. Upon the whole, I may define the posi- 
 tion which I imagine the Caucasians held with respect 
 to the Thracians, both in Asia and Europe, by compar- 
 ing it to that which the Dravidas now hold with respect 
 to the Sanskrit nations in India. The Cappadocians 
 or White Syrians, who divided the Armenians from 
 their kindred, the Phrygians, would have been Semitic 
 invaders from the south at a later period, whom an 
 infusion of Thracian and Caucasian blood may have 
 rendered fairer in complexion than the rest of the 
 Aramaeans. As the Thracians proceeded from Asia 
 towards the west, their language, previously liable to 
 be affected with Caucasian and Semitic elements, would 
 probably have taken up some Celtic, and perhaps some 
 Finnish words, especially in Italy and the Eastern 
 Alps. But its substance and its structure would re- 
 main Armenian ; and such, I believe, the Etruscan 
 language will prove on examination. 
 
26 THE ASIATIC AFFINITIES OF 
 
 CHAPTER II. 
 
 ETRUSCAN SEPULCHRAL INSCRIPTIONS. 
 
 It lias been shewn in the preceding chapter, from the 
 accounts of the ancients, how they made a series of 
 Thracian nations to extend from the Caspian Sea as 
 far as the Carpathians and the sources of the Rhine ; 
 and several linguistic coincidences have been brought 
 forward in confirmation of the truth of these accounts, 
 which there does not seem to be any reason for im- 
 pugning. There is then, as Etruria is included in the 
 Thracian area, a certain antecedent probability that 
 the Etruscan language would be found to belong to 
 the same Aryan family as the Armenian : and we 
 have, besides this, the evidence of Livy, a native of 
 Padua, that the Rhsetian language, which appears 
 from its relics to have been like the Armenian, was 
 Etruscan with a corrupt pronunciation. It will accord- 
 ingly be the object of this and the two following 
 chapters to shew that the Etruscan language was 
 Aryan of the Armenian type ; and the argument will 
 be opened by an examination of the Etruscan epitaphs 
 that contain the often cited avil and ril, words both 
 characterised by one of those I terminations to which 
 I have already called attention. 
 
THE OLD ITALIANS. 27 
 
 There are three words, cetas, annus, and vixit, which 
 continually occur in Latin epitaphs, accompanied by 
 the name and age of the deceased. The case is the 
 same in Etruscan, where the corresponding words are 
 avil, ril, and leine, with line elsewhere. To illustrate 
 the use of these Etruscan words, it will be sufficient to 
 select the following seven of Lanzas epitaphs, in each 
 of which the proper name is omitted, and its place 
 supplied, for convenience of reference, by the number 
 of the epitaph : — 
 
 453. civil xxxiii. 454. avil ril lxv. 
 
 32. avils xxix. 455. ril liii leine. 
 
 10. ril xxi, 456. ril leine lv. 
 
 87. line. 
 
 There are almost enough materials here to deter- 
 mine the family of languages to which the Etruscan 
 belonged, for the respective meanings of avil, ril, and 
 leine, are scarcely to be mistaken. I will, however, as 
 so much is to be learned from these three words 
 alone, proceed to prove what a careful observer might 
 very likely perceive on inspection : and this demonstra- 
 tion seems to be the more requisite, as it was said 
 long ago, and has not yet entirely ceased to be re- 
 peated, that all we know of the Etruscan language is, 
 that avil ril means "vixit annos," though it cannot be 
 said which is the noun, and which the verb.* Even 
 
 * " Of their language, chiefly preserved to us in their sepulchral 
 inscriptions, we know absolutely nothing. The only expression 
 that has been satisfactorily made out is the very common one of 
 ril avil, ' vixit annos.'" Murray's Central Italy, p. 256 (ed. 1807). 
 
28 THE ASIATIC AFFINITIES OF 
 
 Niebuhr was at fault here. "Al dire di lui," writes 
 Micaiij in his A ntichi Popoli Italiani (vol. ir, p. 351)^ "la 
 Bcienza dell' Etrusco sarebbe ristretta all' interpetrazi- 
 one certa di due sole parole : avil kil, vixit annos." 
 
 Yet all that can be pronounced certain, or almost 
 certain, is, not that avil ril means "vixit annos," but 
 that of the two expressions, avil ril and ril leine, one 
 signifies " vixit annos," and the other " anno getatis." 
 Ril therefore, which occurs in both expressions, stands 
 for both anno and annos, and would consequently be 
 the Etruscan word for "year" used without declension, 
 or else contracted to its crude form like ann. and 
 cetat. in Latin. Again, as avil and avils cannot both 
 be rendered " vixit," to say nothing of other obvious 
 reasons, it is plain that avil is cetas or cetat., and 
 that avils is cetatis ; so that avil would be declined 
 like an Aryan noun. So far then from its being 
 certain that avil ril means " vixit annos," it is easy 
 to see that avil ril cannot mean "vixit annos." 
 Finally, as ril is "year," and avil is "age," no 
 sense but " lived" is left for the remaining word 
 leine, which, when accompanied by the number of 
 years of life, is always joined with ril, "years," as 
 vixit is with annos * When it stands alone, as line 
 
 * " Occurrit (leine) in titulis sepulcralibus (ex Volaterris), con- 
 junctum cum voce ril, quod exponitur annos vel annorum." 
 Fabretti, p. 1042. Should it not therefore have been inferred 
 that leine corresponded to "vixit," especially when avils, " Eetatis/' 
 had been rightly interpreted, as it has for some time been in 
 Italy? Of the explanations that have been given for leine, the 
 Latin lene, i. e. leniter, and the Greek \divos, are selected as the 
 most probable in Fabrettis vocabulary. But even if lene be taken 
 
THE OLD ITALIANS. 29 
 
 does twice in Lanzi, the Roman euphemism, " she 
 lived," would be employed for " she is dead." 
 
 The Etruscan words, leine or line, " he lived/' ril, 
 "year," and avil } "age/' may be thus explained : — 
 
 Sanskeit. 
 Root li, "adhere, dwell, live" 
 lay a, " house, dwelling."* 
 lindti, "he dwells." 
 alindt, "he was dwelling." 
 
 Armenian. 
 Root li, "become, be, live" (implied in li-eal, "been"). 
 law ay or loray, "dwelling, home" 
 Lore or Lori, name of two towns, " Ham" 
 lini, " he is." 
 liner, "he was" (imperfect: there is no aorist). 
 
 Etruscan. 
 Lori~um, name of a town.f 
 leine or line, " he lived." 
 
 as a valediction equivalent to " sit tibi terra levis," and Kd'ivos as 
 meaning " tumulus/' how can the resulting explanations of ril 
 leine ly or ril liii leine be more probable than " annos vixit . ." ? 
 There is no reason to infer at starting that the Etruscan, Greek, 
 and Latin languages belonged to one family, but rather, as 
 Niebuhr and Micali affirmed, strong grounds for separating the 
 Etruscan from the other two : for the Greek and Latin, after 
 being put to the torture for a century, have' given no explanation 
 of the commonest Etruscan forms. Proximity need not imply 
 affinity;, and the languages of Etruria and Latium may have 
 been no more nearly allied than are those of England and Wales. 
 
 * Cf. Thracian Ae'jSa, te w6\i$." Botticher's Arica, p. 51. 
 
 f Loro and Lari are two modern towns in Tuscany. 
 
so the asiatic affinities of 
 
 Gaelic. 
 
 Root ra, " go." 
 
 ra-idh, "a quarter of a year/' 
 Sanskrit. 
 Root ri y "go." 
 
 ri-tx\, " a season (of two months)." 
 Armenian. 
 Root rail, " go" (implied in rah-el, "to go"). 
 kath-i7, "to drop, a drop" (root hath). 
 
 Etruscan. 
 r-il, " a year." Compare n-il = nih-il. 
 
 Sanskrit. 
 Root <zv, " grow,, move." 
 
 Armenian. 
 
 lin-iZ or lin-el, " to be" (root li y base lin). 
 
 lin-eli, "that is to be" (future participle) . 
 
 av-e\i, " exceeding, redundant, more." 
 
 yav-et, " more, rather." 
 
 yav-ite&n, "an age." 
 
 az;-ag, " more aged, elder." 
 
 tes-z7, "aspect" (root tes). 
 Etruscan. 
 
 av-il, "age." 
 
 The Armenian av-r, "day, time, age," and dv-t 9 "ring," 
 i. e. " circle, circuity orbit," might also be akin to 
 avil and the Sanskrit av : and the following words 
 may be cited to illustrate the connexion between 
 "going" and "year/' and "season" and " year;" as 
 well as to exhibit still further the relationship of the 
 
THE OLD ITALIANS. 31 
 
 Indians to the Armenians, and of both, in their proper 
 degree, to the reputed ancestors of the Etruscans : — 
 
 Sanskrit. 
 Eoot hi, " go, increase, amplify." 
 hay ana, " year." 
 
 nava, 
 
 garad, " autumn, year ;" whence gdrada, " new." 
 Armenian. 
 Nava-sard, " ancient name of first month :" 
 zardi, "new." 
 
 Lydian. 
 Neo? G-aphts, " New-year."* 
 
 Ossetic. 
 sard, " summer." 
 Something may likewise be said about the forms of 
 the Armenian words adduced to explain leine, avil, 
 and ril. The Armenian imperfect is formed like the 
 imperfect of the Latin possum, so that lin-er, " he was 
 living," corresponds to pot-erat : for the Sanskrit 
 dsit, "he was," = Zend as, = Armenian Sr, = Latin 
 erat.f But, in the imperfect, it is only in the third 
 person singular, and there, it may be, in order to 
 distinguish er, "he was," from e, "he is," that the 
 Armenian retains the s of the Sanskrit root, converted 
 into r as in. Latin. In the Armenian Si, "I was," 
 = Sanskrit &sam, and Sir, "thou wast," = Sanskrit 
 
 # Botticher's Arica, p. 49. The city Sardis might derive its 
 name from the Armenian zardi, "new." Cf. Ebrard on Rev. iii, 1. 
 t The augment is, however, wanting in erat. 
 
o'2 THE ASIATIC AFFINITIES OF 
 
 dsi8 } that letter is omitted; so that we should rather 
 expect in the third person, instead of liner, a form 
 like the Etruscan leine or line, "he lived/' which may, 
 too, be the Sanskrit (a)lin (t), "he was living/' when 
 the augment is dropped, as in Latin, Zend, and Ar- 
 menian, and the final t, as in Armenian, Greek, and 
 Italian. But perhaps, as we shall see eventually, 
 leine is an historic present, = Armenian lini, while 
 line is the imperfect. With respect to av-il and r-il, 
 it should be noticed, that -i7 is the usual termination 
 of the passive or neuter infinitive in Armenian, as in 
 lin-il (or lin-el), "to be." The Etruscan ril, "year," 
 i. e. "course (of the sun)/' would thus signify originally 
 "das Gehen," = Armenian rahel, just as the Armenian 
 kathil signifies " a drop/' as well as " to drop :" and 
 civil, "age," would in like manner signify "das 
 Wachsen." The German leben is another example of 
 an infinitive that is also a noun : and it is remarkable 
 that a single Armenian word lin-il, " leben," should 
 contain the meaning of leine or line, and the form of 
 av-il and r~ih In the Etruscan lein-, too, there is a 
 double affinity to the Sanskrit and Armenian ; for the 
 conjugational n is there, as well as the root li or IL 
 
 In the Armenian lavray or loray, " a home," which 
 explains the signification of the Armenian Lori and 
 the Etruscan _Lon-um (ante, p. 29), the root of dwelling, 
 II or li, seems combined with the Sanskrit vri, "tegere," 
 which occurs, with the termination -an added, in the 
 Armenian tcran, "tent, booth." Compare Yevcma, 
 the Noric Virunum, and the village Vrin in the Grisons. 
 
THE OLD ITALIANS. 33 
 
 Avil or avils 9 "getatis," accompanied by the number 
 of years of life, is found in conjunction with two other 
 words besides ril. One of these words appears in two 
 forms, lupu and lupuhe, the last being one of those -he 
 forms which are so common in Etruscan, as in tur-he 
 or turu-he, "he gives. " In Lanzi we find : — 
 465. lupu. 
 
 463. lupu ami xxm. 
 
 464. lupu avils xvn. 
 and in Fabretti : — 
 
 2100. avils xxxvi lupu. 
 
 2058. avils lx lupuhe. 
 If ami ril corresponds to anno cetatis, and ril leine 
 to vixit annosy the Latin form to which lupu avils cor- 
 responds would most likely be obiit cetatis. Lupu 
 would then mean " he dies," and be a verb belonging 
 to a u conjugation, like the Armenian lizov, "he licks." 
 The root of lup-u is supplied by the Sanskrit lup, 
 "destroy;" lup, "kill, rob;" or by the Polish tup, 
 "booty;" tup-id, "to plunder;" tup-ac, "to split;" 
 or by the Gaelic lobh, " putrefy," and the Irish labha, 
 " corpse." All these words are allied to the Sanskrit 
 ln } == Greek \v-co, and to the Armenian lovz-el, " to 
 loose" We shall meet in another epitaph with lupum, 
 which means "corpse" in the accusative.* 
 
 I said it was only almost certain that avil ril and ril 
 leine meant " anno setatis" and " vixit annos," because 
 it would be possible a priori that lupu avils meant 
 
 j * The Greek and Latin explanations of lupu are \oirds and locus. 
 
 D 
 
3 L THE ASIATIC AFFINITIES OF 
 
 "vixit annos," and ril leino, "setatis obiit." The forms 
 might be sufficient to assure us that, in these four 
 words, ril and avil are the two nouns, and lupu and 
 h 'mi' the two verbs. If, therefore, we can deduce, from 
 the Armenian and Sanskrit, avil, " age," ril, " year," 
 lento, " he lives," and lupu, " he dies," we should have 
 the right meanings for the four words collectively, and 
 might fairly conclude that each one of them was cor- 
 rectly interpreted. Nor would this conclusion be 
 much, if at all, shaken^ because avil with a number is 
 found connected with one more word, as in the follow- 
 ing epitaph, where the proper name is omitted as 
 before (Giorn. Arcad., vol. cxix, p. 325) : — 
 zilachnhe avil S.I. (qu. avils . .) 
 Orioli renders these words conjecturally, " obiit, de- 
 
 positus est, sepultus est (' o simile') aetatis ;" and 
 
 it is not easy to see how a different meaning can be 
 given to them. As, therefore, the sense "obit" 
 (= "obiit") is in all probability anticipated by lupu, 
 zilachnhe ought apparently to mean " sepelitur, info- 
 ditur": and, as the Aryan character of the Etruscan is 
 sufficiently apparent in the forms already interpreted, 
 a root found in many Aryan languages would be an 
 appropriate root for zilachnhe. Such a root appears in 
 the Greek Xa/co?, in the Gaelic lag, " cavum, specus," 
 in the Italian lacca, "fossa, caverna," in the Armenian 
 etag } " fossa," and in the Phrygian lachit, which pro- 
 bably means " fodit." Assuming, then, lach as the 
 Etruscan for "grave," and "is buried" as the meaning 
 of zilachnhe, we might make the following comparisons 
 between Etruscan and Armenian forms: — 
 
the old italians. 35 
 
 Etruscan. 
 
 lack, " a grave.-" 
 
 zi — lack nk e, " he is buried/' 
 
 Armenian. 
 
 akn, "an eye." 
 
 z — akan e, " he eyes." 
 
 ett, "a place. " 
 
 z — etet £, " he places." 
 
 getin, " ground, terre" 
 
 z- — getn e, Ci il terrasse" 
 
 phokh, " a response." 
 
 phokh e> "he exchanges." 
 
 phokh — anak,... " change, lieutenant, vicar." 
 phokh — anak-e, " he exchanges, he succeeds." 
 
 yatth, " great." 
 
 yatth e, " he conquers." 
 
 yatth — anak,... " victory." 
 yatth — anak — e, " he triumphs." 
 The only discrepancy here is, that the Etruscan 
 zilachnke is passive, while the Armenian verbs cited 
 are active ; though yattha.nake, " he triumphs," and 
 phokhanake, "he succeeds," i.e. "he puts himself in 
 the place of another," may be considered as reflective. 
 In Armenian, n implies " self," and k is causative. 
 
 Z is prefixed to nouns and pronouns, as well as to 
 verbs, in Armenian. It distinguishes the objective 
 from the nominative : e.g. sirel zAstovaz, " to love 
 God, amar a Dios" It also marks other cases : as — 
 erthal zoskvoy, " to go for gold" (oski) ; — arkanel 
 znowav, " to put upon him." On the whole, whether 
 
36 THE ASIATIC AFFINITIES OF 
 
 prefixed to verbs, where it is augmentative or deter- 
 minative, or to pronouns and nouns, where it is the 
 latter, its force seems nearly the same as that of eV/.* 
 
 In addition to zilachnke axil . . , lupu avils xvn, and 
 avils lx lupuke, we find in Fabretti (2059), zilachnuke 
 lupuke, "infoditur,moritur," — "he is dead and buried." 
 This shews that zilachnke is not synonymous with lupu 
 and lupuke, and can, therefore, hardly mean anything 
 else than " he is buried" or " interred." If so, and the 
 root be lach, = Armenian etag, " fossa," the manner 
 in which the Armenian enables us to build up zilachnke 
 is very remarkable. As the Armenian phohh gives 
 phokhe and phokhanake, "he exchanges," we should 
 first get etaganake, "fodit(ur);" and, as ett gives 
 zetele, "he places," we should next get zetaganaJve, 
 " mfodit(ur)," = zilachnke or zilachnuke. 
 
 I now come to nine epitaphs of the greatest value, 
 as they contain Etruscan numerals, not in figures, but 
 in ivords. I shall, therefore, give them at length, with 
 their numbers in Fabretti : — 
 
 2104. LarthiKeisi Keises Velus Yelisnas Ravnthus sech 
 avils sas Amke Uples.f 
 
 * The Armenian word zi, = Zend zi, = Sanskrit hi, signifies 
 " for, nam, denn, yap" 
 
 f Amke Uples was probably the person who provided the tomb, 
 or undertook the burial. See 2070 and 2340, where Amke seems 
 the nominative to kepen tenu and Jcisum tame . ., "offers the grave," 
 and "buries the corpse." Kepen may be the accusative of a noun 
 corresponding to the Armenian govb, gen. geb-oy, "a ditch, a 
 cistern ;" and tenu would be equivalent to the Armenian tani, 
 " tenet, tendit." Kisum, as will be shewn subsequently, may be 
 the accusative of Ms-, "a corpse," and tame "buries," appears 
 
THE OLD ITALIANS. 37 
 
 2119. Vipinanas Vel Kla 
 
 nte Ultnas La (r) thai klan 
 avils tivrs sas 
 2033 bis. Yel Leinies Larthial Ruka Arnthialum 
 klan Velusum prumaths avils sesphs 
 lupuke 
 2071. Larth Churchles Arntlial Churchles Thanch- 
 viius(k) Krakial 
 klan avils Memzathrms lupu 
 2070. Arnth Churkles Larthal klan Ramthas Pevtnial 
 vilk Parchis Amke 
 
 Marunuch Spurana kepen tenu avils maclis 
 semphalchls lupu 
 2340. Ramtlm Matulnei seek Markvs Matulm . . . 
 
 puiam Amke Setkres Keis(in)ies kisum tame. . .u 
 Laf . nask Matulnask klalum* ke . s kiklena P.m. 
 a . .. avenke lupum avils (m)aclis mealchlsk 
 Eitvapia me. . . 
 2335a. Larth Arnthal Precus klan 
 Ramth(a)s Apatrual eslz 
 zilachnthasf avils thunesi muvalchls lupu 
 
 2'oQbd. A . . ikne . . eltna turefnesithvas 
 
 avils his muvalchl 
 
 2108. Vipinans Sethre Velthur . . Meklasial Thanchvilu 
 avils his healchfljs 
 
 akin to the Armenian damban, " el sepulchre." Ravnthus ( 2104) 
 should probably be Ramthas, as 2070 and 2340 seem to shew : but 
 I shall not correct the proper names. 
 
 * Klalum, "mcerorem, funera." Greek kAcuw. Armenian lal, 
 " mourning, lamentation." 
 
 t Qu. zilachnuke, " sepelitur." 
 
38 THE ASIATIC AFFINITIES OF 
 
 Iii 2310, we meet with . . avenhe lupum instead of 
 lupu. This form must be explained before proceeding 
 to analyse the numerals. The root of . . avenhe, one of 
 the Etruscan -he forms, like lupuhe, turuke, turhe, and 
 erske, would probably be the Sanskrit av, " servare," 
 which is found in the Armenian ap-aven, " refuge ;" 
 ap- being equivalent to the Sanskrit apa, ap-, and the 
 Greek airo, air-. Apavini means in Armenian, "he 
 takes refuge, he consigns himself;" so that the active 
 apavine would mean " he consigns," and the Etruscan 
 . . avenhe lupum would be rendered " deponit corpus." 
 But . . avenhe might, perhaps, be better explained, 
 still keeping to the same root av, " servare," from the 
 Armenian : — 
 
 avanfdj* " deposit, consignment." 
 avand-el, "to deposit, to give up." 
 avande zhogin, "he gives up the ghost fhogij, 
 he dies;" a singular parallel to the Etruscan 
 . . avenhe lupum, " he gives up the body, he dies." 
 I will notice at a later period the terms of parentage 
 or descent in these epitaphs, hlan, sech, and puiam, the 
 accusative of puia, " filia." Their explanation is not 
 necessary, as that of . . avenhe lupum was, to prepare 
 the way for the consideration of the numerals, to which 
 we will now proceed. From the first three of the 
 epitaphs we get : — 
 
 avils sas 
 avils tivrs sas 
 
 * Compare relva* and tendo; also Armenian spand, " slaughter/' 
 from span, "kill;" and avan, "village" (= Sanskrit avani, "terra"), 
 and sar-avancl, "headland" (sar, "head"). 
 
THE OLD ITALIANS. 39 
 
 avils sesphs lupuke 
 
 " astatis obit." 
 
 I need scarcely cite such forms as avils xxxvi Iwpu and 
 avils lx lupuke, to prove that sesphs, in avils sesphs 
 lupuke, is a numeral : but I must defer its consideration 
 for the present, as Fabretti gives the reading as sesphs 
 in the inscriptions, but as semphs in the vocabulary ; 
 and semph, as will be seen later, appears to be the 
 Etruscan for " seven.-" In amis sas, and avils tivrs sas, 
 sas would be " six," = Sanskrit sas, = Persian sas, = 
 Lithuanian szeszi, = Latin sex, = Greek e£, = Ar- 
 menian ivez, = Afghan sbaz, = Zend khsvas, = Os- 
 setic achsaz: and tivrs would be "thirty/' = Welsh tri- 
 deg, = Latin tri-ginta, = Lithuanian trys-deszimtis, 
 = Sanskrit trim-gat, = Zend thri-gata, = Armenian ere- 
 sovn, = Afghan der-s. This last form is very like the 
 Etruscan, which should, perhaps, be tiers, as E and V 
 fFJ are easily confounded. 
 
 We now pass to the fourth epitaph, which gives : — 
 avils kiemzathrms lupu 
 
 " astatis obit." 
 
 This epitaph belongs, as is seen from the eflfigy, to 
 an old man fuomo vecchioj, and would, therefore, in- 
 volve "fifty," "sixty," "seventy," or even "eighty." 
 If kiemz- be put by the side of tivrs or tiers, " thirty," 
 it will be apparent that kiem may mean " five," and 
 the following table will shew how it ranges with other 
 Aryan forms for that numeral : — 
 Sanskrit. . .pancan 
 Persian... pang 
 Lithuanian. . penlci 
 
40 THE ASIATIC AFFINITIES OF 
 
 C king 
 
 Armenian < . 
 
 . c quinque 
 { quin- 
 Gr&elic.xuig 
 Etruscan... Mem 
 Swedish,., fern 
 Gothic... fimf 
 Welsh... pump 
 
 Greek j 77 ^ 6 
 
 , I 7T6VT6 
 
 Afghan... pinza 
 
 Kiemz- thus means " fifty," = Sanskrit pancagat, = 
 Persian pangah^ Afghan panzus,= Armenian yisovn, 
 = Swedish f emtio, = Gaelic caogad, = Latin quinqna- 
 (jinta, = Bohemian padesat. The termination of Memza- 
 thrms would seemingly involve #A<r-, " three," = Sans- 
 krit £ri, = Zend £/w% = Armenian er and erefchj, = 
 Afghan cKre : and -m- in -thrms might be the sign of 
 the ordinal, as in the Latin pri-m^s and the Lithuanian 
 pir-mas, "fir-st," which Bopp compares with the 
 Sanskrit para-ma, " exiraius, summus." The final s in 
 hiemzathrms would then mark the genitive, and the 
 whole word would signify " (anni) quinquagesimi 
 tertii.-" If so, there are five Aryan characteristics in 
 Mem-za-thr-m-s. 
 
 The sibilants in -* and -z, " -ginta, -kovtcl" would 
 indicate that the Etruscan language did not belong to 
 the same Aryan family as the Greek and Latin, and 
 that it was not Celtic or Teutonic. In the Etruscan 
 
THE OLD ITALIANS. 41 
 
 -5 or -2, the n of the Armenian -sovn or san, " -ginta," 
 is dropped, as it is in the Afghan -s or -s. There is 
 the same omission in the Ossetic das, the Hindustani 
 des, the Persian dah, and the Welsh deg, " ten/' which 
 are all == Sanskrit dagan, Armenian iasn or tasan, 
 Latin decern, Gothic taihun. This n is wanting also in 
 the Sanskrit -gat and the Zend -gata, " -ginta ;" but an 
 additional t suffix is introduced, as in Latin and Greek, 
 of which the Etruscan and Armenian know nothing. 
 The Old Slavonic has not only this t, but retains be- 
 sides, like the Teutonic languages, the da of dagafnj 
 in the Sanskrit -(da)qat, " -ginta. " Thus the Slavonic 
 for " thirty" and "fifty" are tri-desjatj and pjatj-desjatj, 
 which are very unlike the Etruscan tiers and hiem-z. 
 The Lithuanian has trys-deszimtis and penkios-desz- 
 imtis for " thirty" and cc fifty/' The Etruscan problem 
 seems thus nearly reduced to a choice between the 
 Sanskrit and the Armenian ; and if the Sanskrit could 
 be got rid of, the Armenian would then be left alone. 
 Another letter- change, which will be noticed later, 
 may be able to do this. 
 
 The last five epitaphs exhibit these forms :— 
 
 avils Ms muvalchlfsj . 
 
 avils his kealchfljs. 
 . . avenke lupum avils (mjachs mealchls* 
 
 avils machs semphalchlsf lupu. 
 
 * In accordance with the other forms, I drop the final k here, 
 which would probably belong to the following word. Such mis- 
 takes are not uncommon. 
 
 f Thus written in Fabretti's inscriptions, which the other 
 forms shew to be right : in his vocabulary, semphachls. 
 
42 THE ASIATIC AFFINITIES OF 
 
 avils thunesi muvalchls lupu. 
 " setatis obit." 
 
 The age of the deceased is obviously given by the 
 words in italics. What then is the meaning of -alchls 
 or Iclils, which is found in all five of these numerals ? 
 Or rather, what is the meaning of -lchl, for s may be 
 the sign of the genitive or plural ? In all probability, 
 -/c/Jis"-decim," or "-ginta," or "-genti." But it is not 
 likely to be " -ginta," for tiers and hiem-z are " tri- 
 ginta" and " quinqua-ginta." Nor would it be 
 " -deeim," for machs semphalchU is shewn by the effigy 
 on the tomb to be the age of an " old man/' who was 
 certainly out of his teens, unless they were teens of 
 lustres. Lchl would thus appear to be " -genti," = 
 centum. Now the Latin centum and the Sanskrit gata 
 are each 10 X 10, and= decem-decem-tum and dagan- 
 dagan-ta. Lchl is probably a similar form, as may be 
 intimated by the repetition of the I, and = Ich-lch, 
 10 x 10. We have therefore to trace Ich, "ten." 
 
 It is found first in the Lapponic lokke, " ten/' where 
 we have also the form lohhad lohke, "'quod dicit decimum 
 decern, hoc est centum" (Ihre) . The Lapponic lohh-et, 
 " numerare, legere," shews that the fundamental idea 
 in lohke, " ten/' is the same as in Se/ea, which is con- 
 nected with SeUo), = dico, "\ey(o." In fact, lokke is 
 " digit, number/' and is allied to the Lapponic suffix 
 -lokk, " omnis, unusquisque/' from which we may pass 
 to the Armenian loh, " solus, simplex/' and to the 
 Tuschi -loghe or -Ighe, which forms ordinals out of 
 cardinals, as in yethchloghe, "sixth/' from yethch, 
 "six," and qhalghc, "third," from qho, "three. 
 
 )> 
 
THE OLD ITALIANS. 43 
 
 Lch, " ten," may also be explained from the 
 Lithuanian -lika, " -leven, -teen," i.e. "ten," which 
 would = the Polish lik, " number." It cannot thus be 
 said that the Etruscan -Ichl, " -genti," is necessarily of 
 Turanian origin; though, if it were so, it would not be 
 surprising, as the Alpine lawine, "avalanche/' and 
 kass, "glacier/' appear to be Turanian words, and 
 the latter = Lapponic kaisse, " mons altior, plerumque 
 nive tectus." Yet I think Ichl is most likely Turanian; 
 for there were, as will be found, Turanian numerals in 
 Etruscan, and the Aryan for " hundred" in Etruscan 
 seems to be tesnsteis, a reduplication of tesne, " ten," 
 = Armenian tasn. 
 
 All now comes out easily. Me-a-lchls and muv-a- 
 Ichls both signify " one-hundred" or " one-hundredth;" 
 the Etruscan me- and mnv-, " one," corresponding to 
 the two Armenian forms, mi and mov, " one," withme- 
 in me-tasan, " eleven." The connecting vowel a, in 
 me-a-lchls and muv-a-lchls, and in the other similar 
 Etruscan forms, is the same as in Armenian, where we 
 find mi-a-pet, " y^ov-apyo^ ;" mi-a-kin, " having only 
 one wife ;" char-a-chayl, " quadruped ;" char-a-kerp, 
 " quadriform." Ke-a-lchls is " five-hundred " or "five- 
 hundredth ;" for it will soon be seen that "five" is ki, 
 as well as kiem, in Etruscan. We have, too, the Ar- 
 menian yi-sovn, "fifty," as well as King, "fi>e;" and 
 we know that m is elided in Latin between two vowels, 
 as in co-arguo, = cum-arguo. Finally, semph-a-lchls, 
 if semph be right, is " seven-hundred," or " seven- 
 hundredth ;" semph-, " seven," being = Latin septem, 
 
44 THE ASIATIC AFFINITIES OF 
 
 Sanskrit saptan, Slavonic sedmj, Russian sem, Zend 
 haptan, Armenian evthn, eothn, Greek eirrd. The 
 Armenian evthn, " seven/' would explain the Albanian 
 yave and the Rhseto-Romance evna, "week/' just as 
 the Kurdish ahft, " seven/' explains the Kurdish 
 ahftie, "week." The present Albanian word for 
 "seven" is sta(te), from which yave could not be 
 derived. The Armenian for "week'' is eothneak or 
 eavthneaJc. 
 
 The Etruscan " -genti" forms may throw some light 
 on another question. The Aryans use three distinct 
 words for " thousand." The Slavonian and Lithuanian 
 terms are like the German toMsend, as the Armenian 
 hazar is like the Sanskrit sahasra, and the Celtic mile 
 and mil are like the Latin mille. The Etruscan mealchl } 
 "one-hundred/' might lead us to explain mille as 
 "one-thousand/' as if =mi-lch X Ich X Ich contracted ; 
 and it is possible that %l\-iol might be " thousand/' = 
 M% X M% X ^%)j as centum is (decende) centum. The 
 forms of mil and ^iX- are like the Tuschi met, " quot/' 
 which appears derived from me, " qui/' by the addition 
 of t, " number/' = Lapponic lokke, == Etruscan Ich, 
 =: Polish lih. "How many there are" = " what a 
 number there are." 
 
 Nothing remains for explanation but machs, thunesi, 
 and Ms. Now machs semphalchls, as is known by the 
 effigy, is the age of an "old man/' who might have 
 lived nearly sixty years, or about " seven-hundred 
 months." Machs would thus be the genitive or plural 
 of mach, " a month/' == Sanskrit mds, " moon/' — 
 
THE OLD ITALIANS. 45 
 
 Persian malt, = Armenian mah-ih (a diminutive in ih), 
 " the crescent moon." The letter-change in the Etrus- 
 can much, from the Sanskrit sibilant to the Aric 
 aspirate, should not be overlooked ; but in the Ar- 
 menian amis, "month," it is not observed. Finally, 
 as ril is "year," and mach is " month/' thunesi (p. 42) 
 would be " days," or "day's," and is explained by the 
 Armenian tovngean, which is used as the genitive of 
 tiv, "a day," and might be the proper genitive of 
 tovnig, as teslean is of tesil. Lars (2335 a) would be 
 an infant who only reached the age of a hundred days, 
 while Ramtha (2340), who is described as pniam, 
 " filiam," would be a girl of a hundred months old, or 
 in her ninth year when she died. I have not dis- 
 tinguished the two Etruscan characters for s, one of 
 which, that in thunesi, is supposed to correspond to the 
 English sh, which is nearly the Armenian g. 
 
 The Etruscan thvnes(i) and the Armenian tovnig sup- 
 
 pear to be composed of the Armenian tov(oy), the 
 
 genitive of tiv, "a day," — a word which may be allied 
 
 , to thiv, gen. thovoy, " number, year, epoch," — and of 
 
 the Armenian wis, gen. nsi, i.e. nesi, " sign, mark," = 
 
 Hebrew nes. The Armenian combines nis with the 
 
 , pronouns, ays, ayn, " this, that," in the expressions, 
 
 i ays nis, ayn nis, " such a one, 6 Selva ;'' both being 
 
 forms similar to the Etruscan tJtu-nesi, " diei," and to 
 
 what would be a genuine Armenian word, tov-negi, 
 
 with the same signification. And let it be remembered 
 
 that it is not from the resemblance of the Etruscan 
 
 thu-nesi to the Armenian tov-negi that it is interpreted 
 
46 THE ASTATIC AFFINITIES OF 
 
 €t diei." It is from independent argument, not from 
 similarity of sound, that leine, lupu, avil, ril, mach, 
 and thanes, are concluded to mean respectively, 
 " lives," " dies," tf age," " year," " month," and 
 " day," the six terms which we find in Latin epitaphs 
 where the age of the deceased is given. 
 
 Niebuhr noticed that on two occasions the Etruscans 
 made truce with the Romans, once for twenty , and the 
 other time for forty years ; but yet were again at war 
 with them, and apparently without breaking truce, at 
 the end of eighteen and thirty-six years respectively. 
 Niebuhr explains this by saying that the Etruscan year 
 would have contained only ten months. No doubt the 
 explanation is correct, and the truces would have been 
 made for two-hundred and four-hundred months. This 
 shews how the Etruscans were in the habit of reckoning 
 by periods of one-hundred months, each of which periods 
 would have been a kind of double lustre, just as the 
 Latin has bilustris for a period of ten years. Of course 
 such periods could only have been used in epitaphs 
 when the age of the deceased happened to be nearly 
 bilustral ; and this may explain why, in the case of two 
 members of one family (ante, p. 37, epit. 2070, 2071), 
 the age of one of them is defined by the words, avils 
 kiemzathrms lupu, and that of the other by the words, 
 avils machs semphalchls lupu. Larth Churchles lived 
 fifty-three years, ril being understood; and Arnth 
 Churkles seven-hundred months, or about fifty-eight 
 years. But perhaps one-hundred months = eight years.* 
 * An Etruscan week was eight days (Niebuhr). 
 
THE OLD ITALIANS. 47 
 
 Kis may be explained in two different manners. 
 When his muvalchls and his healchls are compared with 
 macks mealchls and machs semphalchls {ante, p. 41), it 
 might be inferred that his = machs, and therefore that 
 hi means " month." If so, we have another Turanian 
 word; for "moon, month," is Ten in Esthonian, and 
 huu in Fin ; while in Georgian " moon, month," is thve, 
 and in Tuschi "white" is hui : all terms probably akin 
 ultimately to the Sanskrit gve(ta), " ' whi(te)."* Other- 
 wise, as there is no equivalent for lupu, " dies," or 
 avenhe lupum, " leaves a corpse," in the two epitaphs 
 which contain his (p. 41), that word might be explained 
 from the Armenian ges, gen. gisi, " a corpse -" and 
 macks, ec months," would be understood, like ril, 
 "years," elsewhere. At any rate, hisum would be 
 rendered "veicpov" in 2340 (ante, p. 37), where it is in 
 apposition to Bamthn. The beginning of that epitaph 
 would m ean : Ramtk am Matu luce pro lem, Marci Matu Ince 
 filiam,, Amy cus a Sethre Ocesennia veicpov sepelit. So, 
 too, in the first line of another epitaph (Fabretti 2339) 
 we read :— 
 
 Larth Keisinis Velus klan hizi zilachnhe — 
 where hizi zilachnhe would be rendered " dies (and) is 
 buried," or " (being) dead is buried/' or " is buried 
 with the dead ;" according as we make hizi an Ar- 
 menian verb like lini (ante, p. 29), or a noun like ovti, 
 " a way/' or a noun like ges, " a corpse." Very pro- 
 bably, his and hisum, are unconnected in sense. 
 
 * Mouna Kea in Hawaii is the " White Mountain." It is the 
 Mont Blanc of the Sandwich Islands. 
 
48 THE ASIATIC AFFINITIES OF 
 
 I now turn to notice the doubtful inscription, avils 
 sesphs lupuke, where the sesphs of the inscription is 
 given as semplis in Fabretti's vocabulary, and there 
 interpreted " seventy/' after the analogy of semph, 
 " seven/' in that form which he gives as semph-alchls 
 in his inscriptions, and as semph-achls in his vocabu- 
 lary. There are thus, in Fabretti, three to one in 
 favour of semph, " seven/' as against sesph, which 
 would have to be referred to the Basque zazpi, " seven." 
 Now, as tiers would be "thirty," semphs or sesphs 
 would rightly be " seventy/' if semph or sesph be 
 " seven." But there is one objection to the interpret- 
 ation. The epitaph, avils sesphs lupuhe or avils semphs 
 lupuke, is annexed to the figure of a young man 
 (giovane), which would seem from the description of 
 the tomb to represent the deceased, though it may 
 not do so. If it does, as the Etruscan ph and th 
 nearly resemble each other, the true reading might 
 possibly be sesths or semths (for semphths), " sixteen" 
 or "seventeen." In this case, -ths would = Welsh 
 -theg, Sanskrit -dagan, Armenian -tasan; the final 
 -an being dropped in -ths, just as in the change from 
 the Armenian -sovn or -san, "-ginta," to da3^Etruscan 
 -.9. On the whole, I should think Fabretti's vocabu- 
 lary and interpretation most likely to be right as to 
 this doubtful word, and that we ought to read semphs, 
 "seventy," rather than anything else. But it is not a 
 word that can be much relied on.* 
 
 * The Etruscan m and s (sh) are liable to be confounded. In 
 the same epitaph, Arnthialum should probably be Arnthialus. 
 
THE OLD ITALIANS. 49 
 
 The Perugiah Inscription, which appears to be a 
 conveyance of land for a burying-ground, may help us 
 to arrive at some other Etruscan numerals. We find 
 there, in two different parts : — 
 
 , ,c/iiemfusle 
 
 and also — — sleletMara 
 
 tezanfnsleri tesnsteis 
 
 rasnes... 
 
 As chiem seems probably = hiem, "five," Jcamdezan 
 may be a number; and if so, would apparently be 
 " fourteen/' = Sanskrit caturdaqan, = Armenian corech- 
 tasan : but in c/mra-sovn, " forty," the Armenian 
 approaches more nearly to the first part of karu-tez&n. 
 The second extract above concludes with — 
 
 tesnsteis 
 rasnes 
 and we have besides — tesne eka velthinathurasth 
 auraheluifesfterasnekei 
 tesnsteis rasneschimthsrj 
 Here rasne would not improbably be the Persian 
 rasan, Armenian arasan, "a cord," Sanskrit rasand, 
 " a girdle," and might mean " fathom." Cf. <T%olvos ; 
 also Gerrm^Mafiery "cord, fathom," faden, "thread, 
 fathom ." Fusle may have been some larger measure 
 than rasne. As we may possibly have sesths, "six- 
 teen," elsewhere, chimths might be " fifteen ;" though 
 this seems inconsistent with karutezan, "fourteen," 
 and is open to other objections. But there would be 
 less reason for doubting that tesne is "ten," = Arme- 
 nian tasn,— Sanskrit and Zend dagan; and if it be, then 
 
 £ 
 
50 THE ASIATIC AFFINITIES OF 
 
 tesnsteis would probably be "ten tens," or "a hundred;" . 
 a formation like Ichl, excepting that tesns-, "tens," 
 has the sign of the plural. The Armenian tasn is 
 declined, the genitive being tasin, and the instru- 
 mental tasamb. The Sanskrit gata, "hundred," would 
 be formed from daca(n)-daca(n)-ta by a process very 
 different from that employed in the formation of 
 tesnsteis. The Armenian for " a hundred" is harivr, 
 which is an entirely different word. 
 
 Kiem, sets, semph, and tesne, have been interpreted 
 " five," " six," " seven/" and " ten," bj^ others before 
 myself, as may be seen in Fabretti's index. I now 
 come to something more remarkable than the exist- 
 ence of Aryan numerals in Etruscan ; a fact which 
 would have been expected in a language that is plainly 
 Aryan. There are numerals in Etruscan which are not 
 Aryan, as Ichls has perhaps intimated already. This 
 important discovery connects itself with what I said 
 in my first chapter. I there noticed that there were 
 in ancient times Iberians, Ligyes, and Tuscans, not 
 only along the Mediterranean from Gibraltar to the 
 Tiber, but likewise all three in the Caucasian regions : 
 also that there are now Toscans in Albania, and TuscJii 
 in the Caucasus; and that the Tuschi word law, "snow," 
 explains the lawine of the Alps. It appears too from 
 Livy (x. 4) that there was a difference between the 
 town and country speech in Etruria. Now on a pair 
 of Etruscan dice (Pabretti, 2552) the first six numbers 
 are given in words : and by comparing the relative 
 positions of these words with the relative positions, on 
 
THE OLD ITALIANS. 51 
 
 a quantity of other ancient dice, of the same numbers 
 expressed by points, Campanari has determined their 
 value to be as follows : — 
 
 I II III IV v VI. 
 
 mach thu zal huth hi sa* 
 Here mach is Aryan, and may be compared with 
 the Armenian miak, " unique, one." Thu and sa are 
 Aryan, though not exclusively, as they are like the 
 Circassian tu and the Basque sei. The Irish for "six" 
 is se. Zal, (C three," is not Aryan. Huth is not Aryan, 
 
 * I do not know whether any exceptions will be taken to my 
 interpretations of machs, Ms, and thunesi (ante, p. 41), because 
 mach, lei, and thu are here found to be Etruscan numerals. But, 
 if machs and Ms be there interpreted as " first" and " fifth," we 
 should then have to understand ril, "year," and therefore to 
 interpret mealchls and muvalchls as "tenth/" Jcealchls as "fiftieth/ 5 
 and semphalchls as "seventieth." This seems to me improbable : 
 and even if it were allowable to interpret muvalchls as " tenth," 
 what are we to make of thunesi in avils thunesi muvalchls lupu, 
 
 "Eetatis (anni for anno) (et) decimi obit ?" Thunesi ought to 
 
 be the genitive of a number, as well as machs and Ms, and would 
 therefore have to signify "eight" or "nine," as all the other 
 numeral places are occupied. Now thu is " two" in Etruscan, and 
 there are languages, of which the Hungarian is one, where "eight" 
 = 4 x 2 ; so that thu-nesi or thu-nes might be " eight," if nesi or 
 nes were " four" : and " four" is negy in Hungarian, and net in 
 Ostiak, two Finnish dialects. Indeed, " eight" in Ostiak is nida, 
 which would = net-da, 4x2. It is thus possible for thu-nesi to be 
 "eight." But, if muva4c7*i-s be "ten," the second I would be 
 non-radical, as Ich, according to analogy, would mean " ten" by 
 itself. Yet this might not be impossible, as the t termination 
 in the Classical, Zend, and Sanskrit " -ginta" and " -genti" forms 
 is non-radical too, and the t termination of the Sanskrit par- 
 ticiple becomes, as will appear later, an I termination in Arme- 
 nian and Etruscan. Or lch-Z might be "ten-/7t," as qk&-lghe is 
 "thir-d" in Tuschi (ante, p. 42). 
 

 J 
 
 ii 
 
 Circassian 
 
 . se 
 
 hi 
 
 Tuschi 
 
 ,zha 
 
 si 
 
 Georgian . . 
 
 .erthi 
 
 ori 
 
 Lazic 
 
 .ar 
 
 zur 
 
 Etruscan .. 
 
 .mach 
 
 thn 
 
 IV 
 
 V 
 
 VI. 
 
 ptVe 
 
 tfchu 
 
 chi. 
 
 dhew 
 
 phchi 
 
 yethch. 
 
 otKkhi 
 
 Jchuthi 
 
 ekhvsi. 
 
 otkh 
 
 khut 
 
 as. 
 
 huth 
 
 M 
 
 sa. 
 
 DZ THE ASIATIC AFFINITIES OF 
 
 and both huth and Tci seem Caucasian. This may 
 appear by comparing these six Etruscan numerals with 
 the same numerals in Circassian, Tuschi, Georgian, 
 and Lazic : — 
 
 in 
 
 si 
 
 qho 
 
 sami 
 
 gum 
 
 zal 
 
 In th-khu-meii, "fifteen/' the Georgian khuthi, "five" 
 is contracted to khu* The Etruscan hi, "five/' might 
 be obtained from the Georgian kh(uth)i, or the Tuschi 
 [ph)chi; or nearly from the Circassian (t')chu; or 
 from the Lesgi chewa and yku, which mean " five" in 
 two different dialects; or from the Abkhasian chu(ba), 
 " five/' where ba is a suffix for all numerals from two 
 to ten, so that chu would be the number " five." The 
 Etruscan huth, "four/' would probably not be the 
 oth of the Georgian othkhi, "four/ 5 but would be re- 
 lated to the Georgian khuthi, " five/' and the Lazic 
 khut, nearly as the Georgian o-thkhi, "four," and the 
 Tuschi ye-thch, " six," are to the Circassian Vchu, 
 "five." Eor the Georgian -thkhi in o-thkhi, and the 
 Tuschi -thch in ye-thch, are both apparently = Cir- 
 cassian i chu, "five," while the Tuschi ye in ye-thch, 
 and the Georgian o in o -thkhi, may both be explained 
 from the Ossetic yu, yen, or yeue, " one." That is to 
 
 * Th- is "ten/' and -rneti is like the Ostiak ordinal sign -met. 
 
THE OLD ITALIANS. 53 
 
 say, the Georgian o-thkhi, "four/' is " one from five/' 
 iv ; and the Tuschi ye-thch, " six/' is " one to five/' 
 vi. So, in the Georgian ekh-vsi, " six," -vsi would 
 probably = Fin wiisi, "five," while ekh- is "one," 
 = Sanskrit eka, == Persian 7/&Z;_, = Abkhasian afra, 
 = Hungarian egy, === Fin ?/&, which becomes t/7?<- in 
 yh-deksa, "nine," i. e. "one from ten," ix. If now the 
 ekh-, " one," of the Georgian ekh-vsi, " six," be pre- 
 fixed to the Georgian khuthi, " five," we could obtain 
 for " one from five," iv, the form ekh-khuthi, which 
 might be changed, by incorporation and contraction, 
 into the Etruscan huth, " four/' in which a %, imply- 
 ing " one," is lost at the beginning of the word, as it 
 would really be also in the Etruscan sa, "six," as 
 well as in the Sanskrit sas. The following tabular 
 view may present the argument more clearly : — 
 
 Etruscan .. 
 
 -.(%)— 
 
 -sa, 
 
 vi (= one + five). 
 
 Abkhasian 
 
 ... 
 
 sva, 
 
 x (= fives). 
 
 Ossetic 
 
 ....ach — 
 
 • saz, 
 
 vi (= one +five*). 
 
 Zend 
 
 ...M 
 
 -svas, 
 
 VI. 
 
 Sanskrit .. 
 
 ( eka, . 
 
 . . . aka, . 
 
 f yu 
 
 -sas, 
 
 VI. 
 .. I. 
 
 Abkhasian 
 
 
 ..I. 
 
 
 
 . . I. 
 
 Ossetic 
 
 i y w, . . • , 
 
 * ( yen, . 
 
 
 .. I. 
 
 Circassian 
 
 Vcliu, 
 
 V. 
 
 Tuschi..... 
 
 ....ye 
 
 - tlich, 
 
 VI. 
 
 # That saz, etc., probably = " five/' will be shewn in the last 
 chapter. The true value is seen in the Basiiui zazpi, "seven," 
 where -pi = bi s " two." 
 
THE ASIATIC AFFINITIES OF 
 
 Georgian . 
 
 Lazic 
 
 Etruscan ,. 
 
 Lesgi — 
 Circassian 
 Tuschi — 
 Abkliasian 
 Circassian 
 
 r 
 
 |(x)- 
 
 (%)- 
 (%)- 
 
 - thhhi, iv. 
 hhuthi, v. 
 
 - tick, IV. 
 Mint, v. 
 
 —hath, iv. 
 hi, v. 
 
 chew a, v. 
 -chi, vi. 
 
 phchi,* v. 
 -jphsi, iv. 
 jpse, x ( = fives). 
 
 It will be perceived from the forms at tlie head of 
 this table how a guttural aspirate is lost before s. 
 But such an aspirate would be still more likely to be 
 lost when it preceded another aspirate, as in that 
 case the fusion of the two aspirates into one would be 
 almost sure to take place. And so, by the loss of the 
 prefixed aspirate implying I, the Etruscan hath, iv, 
 may = the Georgian hhuthi and the Lazic hhat, v, 
 nearly as the Abkliasian phsi, iv, = Tuschi phchi, v, 
 and as the Circassian chi, vi, = Etruscan hi, v. As 
 Octavias is written JJhtave in Etruscan, the Etruscan 
 h would be guttural, like the Hebrew cheth which it 
 represents. The explanation of the Etruscan zed, 
 " three," must be reserved to the fifth chapter, where 
 the subject of numerals can be more fully discussed : 
 at present I can only exhibit it in connexion with 
 some other numerals for " three," to which it appears 
 wholly or partially allied : — 
 
 # Cf. Georgian johekhi, S( foot." 
 
THE OLD ITALIANS. 55 
 
 ._.. ., f Fin ho-lmi 
 
 Finnish A . . . • 
 
 ( byname Icu-jm* 
 
 " three" = 
 
 ( Lazic ffu-m 
 
 Caucasian ... < Mingrelian ...su-mi I "2 from 5," 
 
 \ Georgian sa-mi n y # 
 
 Etruscan za-l 
 
 Javanese td-lu 
 
 Another representative of " five/' I or l-m, is intro- 
 duced here, and may also be found in the Circassian 
 pt-l'e, iv, and Fie, vn.f Such terms for " five," as 
 will be shewn in the sequel, are all different words 
 for " hand" or " foot" : the Cornish lau, the Gaelic 
 lamh, the Turkish el, and the Malay lima, all signify 
 "hand." It is rather strange that the Etruscans should 
 have used two languages for the numbers on their 
 dice : but we do something of the same kind in England, 
 as we usually employ French names for the numbers 
 on dice, but sometimes English ; while in the case of 
 cards we begin with French, but soon pass into English. 
 If the language of the Rasenae be represented by 
 Anglo-French, and that of the Pelasgians by English, 
 then the six numbers on the Etruscan dice would be : 
 ace, deuce (or two), three, four, five, six. 8a might be 
 sice, but I think it more likely to be six, as sas would 
 be the Etruscan representative of sice. Yet, as sa is 
 
 * In speaking", we drop I before m in holm, and in halm y calm, 
 psalm: but we retain it in elm, helm, and whelm, and also in film. 
 
 f In Basque, bat is i, hi is n, and lau is iv, perhaps rightly v. 
 If so, the two Circassian forms are easily explained from the 
 Basque, as "one from five," and " two to five." 
 
Ob THE ASIATIC AFFINITIES OF 
 
 " six" sas might be " sixty" in Etruscan, if we had 
 not avils tiers sas as well as avils sets. But how 
 would the Etruscans have expressed " sixty/' which, 
 if sas were "six" ought according to analogy to be 
 scts-s or sas-z ? The Armenian gets over the difficulty 
 by changing z into th. Thus "six" is wez; " sixteen/' 
 westasan ; and " sixty," wathsovn* 
 
 I have spoken of the Etruscan numerals, zal, huth, 
 hij and sa, as being Pelasgian. Now that there were 
 Pelasgians in Etruria is so commonly known that one 
 quotation will suffice. The writer who is called Scymnus 
 Chius says : — 
 
 M.era ttjv AcyoartKrjv UeXaajol 8' elalv, oi 
 Ylporepov KaroacrjaavTes etc t% c E\\aSo9, 
 K.OLvr]v Be TvppTjvoLat, %copav vefiojuevoL. 
 To what element in the population of Etruria, unless 
 to the Pelasgians, can we attribute the numerals in 
 question ? We know the Umbrian language to be an 
 Aryan dialect resembling the Latin, and we meet con- 
 tinually with tre and tref, " three/' in the Eugubine 
 Tables. The Tyrrhenians, again, as distinguished above 
 from the Pelasgians, are there derived, according to 
 the common tradition, from Lydia, and would have 
 spoken that Aryan language of the Armenian type 
 which we have found predominant in Etruria. The 
 conclusion seems almost inevitable. We learn from 
 Herodotus that the Pelasgians whom he knew spoke a 
 barbarous dialect. It would now appear that it was 
 not even Aryan; although, if the numerals in question 
 be considered as Caucasian, and the origin of the 
 
THE OLD ITALIANS. 57 
 
 Etruscans from Armenia be admitted, it might be 
 urged as possible that the Etruscans borrowed them 
 when there from the Caucasians, who would have 
 possessed Armenia before the Aryans came thither 
 out of Media. I should, however, for my own part, 
 entirely reject such a solution, especially as the numerals 
 are not the only indication of the presence of Caucasians 
 in the South of Europe. 
 
 The following list will comprise such numerals as 
 have been detected in Etruscan : — 
 i. mack, me-, muv- 
 II. thu 
 in. zal, tlir- 
 IV. huthj har- (?) 
 v. hi, he-, Idem-, chiem (?) 
 vi. sa, sas 
 vn. se(m)ph 
 
 x. tesne 
 xiv. harutezan (?) 
 xv. chimths (?) 
 xxx. ti(e)rs. 
 L. hiemz- 
 liii. hiemzathr- 
 lxx, se(m)phs (?) 
 
 c. tesnsteis, mealchl, muvalchl 
 
 d. healcJil 
 
 DCC. se(m)phalcld 
 We see here an Aryan language which contains some 
 Turanian words, as dozen and century are contained in 
 the English language. The explanation has just been 
 
58 
 
 THE ASIATIC AFFINITIES OF 
 
 suggested. The Basence, or Tyrrhenians of Etruria, 
 were Aryans of the Thracian stock, while the Tuscans, 
 who were the Pelasgians or' "Aborigines" (ante, p. 11) 
 of the country, were Turanians, and probably Cau- 
 casians. It seems likely, however, that the blood of 
 the Etruscans, and of the Thracians in general, was 
 more Turanian than their language, as the primitive 
 Caucasian population on which they intruded would 
 have been incorporated with them into one people. 
 The language of the Etruscan nobles may have been at 
 first Thracian, with some Caucasian accretions acquired 
 on the way from Armenia to Etruria, while the lan- 
 guage of the Tuscan commonalty may have been 
 Caucasian, perhaps with a little Umbrian also. The 
 languages would probably in time have coalesced, as 
 Anglo-Saxon, Danish, and Norman-French did into 
 English ; but in Etruria it was the language of the 
 aristocracy which prevailed. The Etruscan language 
 is Thracian, while the English is German.* 
 
 * Mr. Fergusson has inferred, on architectural and other grounds 
 distinct from language, the Turanian character of the Pelasgians 
 and Etruscans. I must come over to his opinion, as far as the 
 Pelasgians are concerned. When I extended the Caucasians into 
 Europe nine years ago, it was not among them, but among the 
 Thracians, that I was inclined to include the Pelasgians. One 
 result of the Thracian invasion of Europe by the Hellespont and 
 the Bosphorus would probably have been a concentration of Pe- 
 lasgians in Greece. To this predominant Pelasgian or Caucasian 
 element in their blood the Greeks may have been greatly in- 
 debted for their beauty of form : nor would that form have under- 
 gone much alteration from what it was at first, if the Thracian 
 element in the Greek population consisted principally of maritime 
 settlers, Leleges, from Asia Minor, and the Italian element of 
 
THE OLD ITALIANS. 
 
 59 
 
 In the epitaphs which gave us the Etruscan 
 .numerals, we met with klan,puiam, and sech, words all 
 indicating parentage or descent. Klan, which corre- 
 sponds to natus in a bilingual inscription, seems con- 
 nected with the Gaelic clann, " offspring, descendants" ; 
 the Welsh plan, "a scion," and plant, "a son ;" the 
 Latin plant a; and the Greek /c\a?, icKdSos, k\cov, 
 Kkdw, <j>Xd(o, rXdo, ; words which may be allied to the 
 Armenian khl-el, "to tear away, to root up (deplanter) ," 
 as glan, the Armenian for " cylinder," is derived from 
 gl-el, "to roll." But another interpretation is given 
 by K. O. Muller {Die Etrusker, vol. i, p. 446). He 
 compares the two inscriptions on the same monument : 
 La. Venete La. Lethial* etera 
 Se. Venete La. Lethial klan 
 and observes : " If etera be taken to mean ' other, 
 second/ klan must be 'first, firstborn.'" Etera is thus 
 ' compared with the Greek erepo?, = Armenian Star. Dr. 
 Donaldson argues in the same manner (Varronianus, p. 
 
 maritime settlers, Hellenes, from Messapia and its neighbourhood. 
 In blood, though not in language, the Greeks and Lycians may 
 have been very nearly allied, like the people of Cornwall and 
 Brittany. Gauls have become French in like manner. 
 
 Of the Etruscan words mentioned by ancient writers, one is 
 apparently Turanian. This word is damnus, "ftnros," which seems 
 allied to the Lapponic tamp, "equus," the Fin tamma, "equa, 
 the Armenian zambik, "equa, jumeni," the Basque samaria, «ju- 
 mentum vectorium, eaballeria," the Albanian samaras, "junien- 
 tum" the Pehlvi djemna, "camel," and the Mantschu temen, 
 " camel " The Armenian has taken up some Turanian words. 
 
 * Lethe, a man's name, and Lethi, a woman's name, are m 
 Fabretti (1<>39). 
 
 
CO 
 
 THE ASTATIC AFFINITIES OF 
 
 171): "If then etera means, as is most probable, the 
 second of a family, Man must mean Hie first or head of 
 the family." This might bring us to the Armenian gbvkh 
 (= Polish gtowa), "head, summit, the first rank;" 
 gVchan zovkn, "chub" (lit. "head-fish"); glkhani, "the 
 chief persons in a city, the nobility." As there exists 
 rain by the side of regen, and wain by the side of 
 waggon and wacjen, the aspirate kh might be dropped 
 in glkhan. 
 
 Another group of words is presented by the following 
 epitaphs in Lanzi : — 
 
 191. mi K&\a,iru fuius 
 315. Lth. Marikane via 
 
 310. Larthi Vetus Klaukes puia 
 
 311. Arnth Vipis Serturis puiak . . . . 
 123. Anes Kaes puil (t)hui . . . . 
 
 The Latin films and the Greek vio<s would be allied to 
 these terms, and also the Gaelic fuil « blood, family, 
 tribe, kindred." In Finnish dialects we find the Hun- 
 garian fiu, " son," the Syrianic pi, " son," and the 
 Esthonian poia, " son." The Wallachian is very com- 
 plete here ; for it gives us puiu, " what is young," fiiu, 
 " son ,"fia, " daughter," sm&fiika, « Ovydrpcov, tochter- 
 lein." I find no such words in Armenian. The next 
 expression appears to be both Finnish and Armenian : 
 
 63. Larthia Kaia Huzetnas Arnthalisa Kafatl sak 
 
 37. Titi Velimnias Akril sek 
 471. Ramthn Matulnei sech .... 
 Lapponic sakko, "proles." Armenian zag-il, "to be 
 born or derived;" zag-el, "to produce young;" zag, 
 
THE OLD ITALIANS. 61 
 
 :c a young bird (z/eocrcro?)/' = Albanian zok. The Ar- 
 menian zag and the Albanian zok mean also " a small 
 bird/' and are both employed to render " sparrow'' 
 (Luke x, (5). 
 
 The next expressions are of peculiar interest on 
 account of the monuments where they are found, as 
 they occur in the celebrated " valley of tombs" at 
 Castel d'Asso, as well as at Norchia and Toscanella, 
 and elsewhere. But their philological importance is of 
 a still higher character, as they seem to dispose of the 
 claims of the Sanskrit to include the Etruscan in its 
 own peculiar division of the Aryan languages, while 
 the claims of the Armenian stand the test, and are 
 confirmed by the Etruscan forms. One of these forms 
 is eka sntJii (accompanied by a proper name in the 
 nominative) ; an expression which Migiiarini con- 
 jectures to mean "hie situs est/' or " questa e la 
 tomba/ 5 and which must, indeed, have some similar 
 signification. The other form is eka suthi nesl ; and 
 one example of it is eka suthi nesl Tetnie, which I 
 should render, "here is buried the dead Titinia, hie 
 eonditur mortua (necata) Titinia;" considering eka 
 suthi nesl to be now represented in Armenian by aha 
 sovzani nekheal, " ecce sese condit putrefactus/' which 
 does not, however, exhibit the affinity which the Ar- 
 menian bears to the Etruscan so clearly as it may be 
 brought out by further consideration. 
 
 The letter-changes in different Aryan languages 
 should be noticed here. Now we have in the first 
 place — Sanskrit hrid = English heart = German herz 
 
G2 THE ASIATIC AFFINITIES OF 
 
 == Greek xap8- = Latin cord- = Gaelic cridhe = 
 Welsh craidd = Armenian sirt = Lithuanian szird- = 
 Polish serce ; and in the next place — Sanskrit gvan = 
 English hound = German hund = Greek /cuz/- = Latin 
 can- = Gaelic ct& = Welsh ci = Armenian som =J 
 Lithuanian szim-.* From this it may be seen that, 
 while German languages retain the Sanskrit h, and 
 convert the Sanskrit g into h } and while Celtic and 
 Classical languages convert both into h 3 = c, (and 
 sometimes into g), Thracian and Sarmatian languages 
 represent both by sibilants. Now let us apply this 
 principle to deal with the roots of — 
 suth-i nes-l 
 
 " cond-itur nec-atus." 
 If these two Etruscan words are rightly translated into 
 Latin (supposing necatus = mortuus), then the following 
 comparisons will shew, by the aid of the letter-changes, 
 that the roots of the Etruscan and Latin are the same 
 in both words, and that the Etruscan is like the Ar- 
 menian in all respects. For we should have : — 
 
 c , . , f hud. " coacervare, submergi/ 
 banskrit-< L \ « „ ° 
 
 ( hund, coacervare. 
 
 J Welsh . . .cudd-, cc to hide." craidd 
 
 \ Latin ...cond-, cc to hide, to bury." cord- 
 
 {Lithuanian .szut-, " acervus." szird- 
 
 Armenian . sovz-, " to hide, to submerge." sirt 
 
 Etruscan . . suth-, " to bury." 
 
 # Botticher defines the Aric, or Thracian and Persian lan- 
 guages, as differing from Indian, German, and Latin, by using h 
 for 8 S z for h, and s for q . 
 
 Cf. hrid 
 
TEE OLD ITALIANS. 63 
 
 Sanskrit. nag, "perire, mori." Cf. dag-sm., "ten.* 5 
 
 $ Latin . nee-, " kill." • dec-em. 
 
 { Greek. vex-, " die." Se/c-a. 
 
 ( Armenian. nas, " a coffin." tas-n. 
 
 \ Etruscan . nes, " die." tes-ne. 
 
 Anglo-Saxon. na,ne," corpse." ty-®-, ec zeh-n." 
 
 Similar letter-changes occur in the Sanskrit spag, 
 "perficere," = Armenian spas, "function, service/' 
 jgpas-el, " to serve, to observe, to watch (Spier)"} 
 = Latin spec-, = German spali-en, = English spy. 
 The Armenian nekh-, " putrescere," found above in 
 nekh-eal, "putrefactus," would not represent the Sans- 
 krit nag, " mori," so much as the cognate Sanskrit 
 nalzk, " necare, destruere." There is a corresponding 
 pair of forms in the Armenian dovstr and dohht, "daugh- 
 ter," = Sanskrit duhitri, Zend duglidar, Persian dokh- 
 ter, dohht, Lithuanian diikte, Slavonic dilsti. The ter- 
 mination of the Etruscan nes-l, = Latin nec-atus, is 
 found in the Armenian nekh-eal. The Slavonic re- 
 sembles the Armenian in having the Sanskrit and 
 Latin t of the preterite participle converted into the 
 weaker I; but the Lithuanian retains the t unchanged. 
 Some Indian dialects exhibit I in the place of t, like 
 the Armenian and Slavonic. In the preterite participle 
 of the Sanskrit nag, the g becomes s; and nasta, 
 "perditus," = Etruscan nesl, "mortuus," which re- 
 sembles closely in form the Slavonic nes-V, "having 
 borne." 
 
 We have considered the roots of the Etruscan suth-i 
 nes-l, as well as the form of nes-l, and have found 
 
64 THE ASIATIC AFFINITIES OF 
 
 tliem all three in the Armenian language. It only 
 remains to notice the form of suth-i, " conditur." 
 Now the general characteristic of the Armenian passive 
 is /, which is equivalent to that of the Sanskrit passive, 
 ya. Thus, in Armenian, "he loves" is sir-3, while 
 "he is loved' 5 is sir-i, which is just the form of suth-i. 
 The corresponding Sanskrit form would be -yate ; the 
 Etruscan and Armenian converting the Sanskrit ya 
 into i, and dropping the t of the third person, with 
 the vowel that follows it. In the aorist, however, the 
 Armenian employs the a, not the y(i) of the Sanskrit 
 ya, to form the passive: thus sirez-i is "I loved," 
 and sirez-ay is " I was loved." Compare tlOt^^l and 
 Tt0€fjL-at, tvttt€{t)-i and TVTTTer-ai* 
 
 The Ehaato-Romance sali/pp, " locust," = Sanskrit 
 galabha, "locust," = Armenian sat ap, "gliding, quick," 
 implies that the ancient Ehsetian was like the Arme- 
 nian and Etruscan in one of its letter-changes. The 
 Greek and Latin give JceU for the Sanskrit gal, "run." 
 The Khseto-Eomance as-oula, "kid" (ante, p. 7), if 
 
 * The characteristic of the Latin passive, r, is Celtic. Thus we 
 find in Zeuss (Grammatica Celtica) these Irish forms:— predch-idir, 
 "■predic-&t\iY" — consuidig-ther, "compon-itur" — tuc-atar, "intellig- 
 untur" — tomn-ib-ther, ( ' cogit-ab-itur" — and prom-f-idir, " prob-ab- 
 itur." In the last two examples, the characteristic b of the Latin 
 future appears also, as again in the Irish car-tt&, " am abo," and 
 the rest of the persons : cair-/e, cair-/ecfc, car -/am, caHr-fid, CELV-fat. 
 In the Welsh par-ass^i, "effeeerit," and agysg-assei, "dormiverit," 
 we meet with forms similar to sun-asset : and in chlyw-yssynt, 
 " Sbud-ivissent, aM&-issent" the Latin form with the root of kXv-v, 
 in-cly-tus, and cZi-ens. The oldest Celtic writings, from which 
 forms like these are taken, reach back about as far as 800 a.d. 
 
THE OLD ITALIANS. * 65 
 
 allied to aly- and the Armenian ayz, would mark a 
 similar peculiarity. 
 
 If the Etruscan suth- = Sanskrit hud, " coacervare, 
 submergi," = Armenian sovz-, "hide, submerge/' 
 then the Aryan family to which the Etruscan belongs 
 would not be Sanskrit, but Armenian, as it is between 
 these two families that our choice would lie. Yet, as 
 the question seems to be reduced to this point, it may 
 be as well to notice another distinction between the 
 Sanskrit on one side, and the Armenian and the 
 Etruscan on the other. The Sanskrit names for a 
 town are pur a, nagara, and pattana, the -poor, -nag ore, 
 and -patam of our present India. But Etruria exhibits 
 none of these names (though pattana resembles the 
 Venetian Patavium), while the corresponding Arme- 
 nian term, sen, gen. sini, does appear there, and may 
 indeed serve to trace the Etruscan route from the 
 supposed " Primeval Country" of the Aryans at the 
 sources of the Oxus and the Jaxartes, up to the Po 
 and the Tiber. The following names, which are all 
 taken from one author, Ptolemy, may be sufficient to 
 exemplify this ; and it will be observed that even the 
 Armenian vowel-change from sen, " village, habita- 
 tion," to its genitive sini, and sin-el, "to build," is 
 not without significance :— 
 
 Among "the mountain-towns of the Sogdians along 
 the Jaxartes" is — 
 
 Xo\/3r)-<T(va* Sogdiana. 
 
 * Ptolemy mentions two Armenian towns called XoAova, and 
 another called XoKov-dra. Cf. Armenian holow, " a round/' 
 
Gi3 THE ASIATIC AFFINITIES OF 
 
 Siva or iLtjva Margiana. 
 
 Sivd/ca* Hyrcania. 
 
 Sava'k Media. 
 
 ^dva ) A 
 
 v / ^ Armenia. 
 
 Zlvis ) 
 
 Savk Phrygia. 
 
 'OTTi-atva or ^OirUariva^ Thrace. 
 
 %evia Illyria. 
 
 Kai-aatva% Cisalpine Gaul. 
 
 A , . M [ Etruria. 
 
 Uvo\-(TLVLOV ) 
 
 Fel-sina, the Etruscan name of Bononia (Bologna), 
 and Sar-sma-, may be added from other authorities. 
 The Lithuanian here touches the Thracian again ; for 
 in Lithuanian sena means " wall/'' and senys, " build- 
 ing.^ The root is also used in Georgian, and Sinope 
 looks as if it were the Georgian senoba, " building." 
 Cf. iifipfa (ante, p. 21). 
 
 Pliny writes, (H. N. } vi. 31) : " Oritur (Tigris) in 
 regione Armenise majoris, fonte conspicuo in planitie. 
 Loco nomen Elegosine est." This is a compound 
 name like Volsinhim, Felsina, and tfarsina. The an- 
 cients mention two places in Armenia called Elegia, 
 and one in Noricum called Elegium. Pliny writes 
 again (xvi. 66) : " Est et obliqua arundo . . . vocatur a 
 quibusdam elegia" One could wish that he had named 
 
 # Armenian diminutives are formed in -ah. 
 f Compare bir- with the Armenian oph, " trench." 
 X Cf. Armenian kay, = Sanskrit kaya, " dooms." 
 § Sena, now Siena. 
 
THE OLD ITALIANS. 67 
 
 these quidam, for elegia would be the Armenian etegn, 
 "a reed/' which appears again in the Phrygian eXvfios, 
 " avXos,"* as well as in the Greek eXeyos, a word 
 probably borrowed by the Ionians from the Lydians. 
 It is said on the Arundel Marble .: r/ Tayvi$ 6 <J>pu£ 
 avXov? 7rpcoro? rjvpev; and by Pausanias (x. 7) : 'EXeyeia 
 teal Oprfvoi TTpocraSo/Jieva to?9 avXois. 
 
 One use that the Armenians made of reeds was, 
 according to Xenophon, to suck up beer or barley- 
 wine through them. This practice the Armenians had 
 in common with the Phrygians and Thracians ; a fact 
 mentioned in his History of Greece by Mr. Grote, who 
 adds, with a just appreciation of national relationship: 
 " The similarity of Armenian customs to those of the 
 Thracians and Phrygians is not surprising." Ncenice, 
 like iXeyeca, were sung to the flute, which is called in 
 Armenian etegnaphot, "reed-trumpet." Thus Cicero 
 says (De Leg., \\, 24) : " Cantus ad Ubicinem, cui 
 nomen ncenice" Compare vnviarov or vcvr/aTos, " ^pvyiov 
 yuiAo?/' on which word Botticher observes : " Ncenia 
 Romanorum in inentem venit, et radix nu, 'laudare.'" 
 This Sanskrit root is found in the Armenian nov-ag., 
 " a song/' and nov-al, " to mew" ; while vrj- or nce- 
 may be referred to the Persian nay, " flute/' — Sans- 
 krit nada, " arundinis species/' = Armenian net, 
 " sagitta/' i. e. " calamus." Thus ncenia seems the 
 nay -mi, "flute- song/' just like eXeyos. The Armenian 
 word for " lute/' win, is the Sanskrit vina, " lute" ; 
 
 # Botticher's Arica, p. 34. 
 

 08 THE ASIATIC AFFINITIES OE 
 
 and the Armenian dinar, "lyre," is obviously the 
 Greek Kivvpa and the Hebrew Mnnor. We may see by 
 these instances what the comparison of languages 
 has exhibited all along from the first (ante, p. 7), 
 how Armenia was connected with India on one side, 
 and still more closely with Asia Minor, Thrace, and 
 Italy, on the other. 
 
 In addition to Lorium (ante, p. 29), and Sena, 
 Fehina, and Yolsinium, or Yolsinii, other Etruscan 
 names of towns may be explained from the Armenian : — 
 Veil, i(j> vtyrfkov aKoirekov, from weh, " high" — Void, 
 oy v O\klov, from ovthh, "a ravine" — Hasta, from hast, 
 " strong" — Blera, from blovr, " a hill," and blrak, " a 
 hillock" — Aharna, from aJcarn, "a castle" — and Nepete, 
 from Mount Npat (Niphates), and npatah, " object, 
 mark." Compare cr/coirta and a/coireXo^. 
 
 I return from this rather long digression, to complete 
 the examination of Etruscan sepulchral forms of ex- 
 pression. 
 
 An Etruscan word for " tomb" appears to be tular. 
 Tularu is found in the Perugian Inscription, and 
 Lanzi supplies these four epitaphs : — 
 
 457. hilar Easnal 
 
 458. tular Hilar .... 
 
 460. tular Svuriu Au. Papsinasl .... 
 
 461. Tetvntertular. 
 
 Tular may be interpreted " tumulus," and thus con- 
 nected with the Greek tv\t], rvXapos, and the Gaelic 
 tula, "hillock"; or else be explained from the Arme- 
 nian that, "tellus," that el, "to bury," thatar, "an 
 
55 
 
 » 
 
 THE OLD ITALIANS. 69 
 
 earthen vessel." Lanzi supposed hilar = to ollarium. 
 No Etruscan phrase has yet been noticed which 
 expresses the sentiment of the Greek (juveias x^? lv or 
 fjLvrjfjL7]<; eveicev, or of our English "in memory of." 
 But I believe that we have one in the following epi- 
 taphs in Lanzi : — 
 
 76. Thutnei thui 
 80. Laris Vete thui 
 313. Thui Larth Petrni Larthalisa. 
 Thui might be interpreted "nominatur, memoratur," 
 by the aid of these Armenian analogies : — 
 thiv (gen. thovoy), "numerus." 
 thovel, "numerare. 5 
 thove, "numerat." 
 thovi, "numeratur/ 
 thovich (plural form ofthovi), "" sententia." 
 The next epitaph (Lanzi, No. 86) would thus imply 
 that the person named is commemorated, and lived (i. e. 
 is dead) : — 
 
 Larth Vete Arnthalisa thui. Larth Vete line* 
 
 Thovi, it will be seen, being implied in thovich, is 
 
 both a noun and a verb in Armenian; and in like 
 
 manner the Etruscan suthi would probably signify not 
 
 only "is buried/' but also "grave, tomb." For the 
 
 following inscription is given by Lanzi (vol. ii, p. 562) : — 
 
 mi suthi Larthial Muthikus. 
 
 (C I (am) the tomb of Muthicus the son of Lartia."f 
 
 # Why should the Latin line, " unge," be considered a probable 
 explanation of this Etruscan line ? 
 
 f We know from Herodotus that the Lydian Myrsil(us) meant 
 "the son of Myrs(us)" 
 
70 THE ASIATIC AFFINITIES OF 
 
 It might, however, signify : " Mutliicus the son of 
 Lartia is buried in me." The stone containing this 
 inscription was discovered at Busca, between Cuneo 
 and Saluzzo, in the country of the ancient Vagienni ; 
 a remarkable locality for an Etruscan epitaph, though 
 it might imply no more than a Greek inscription 
 would do at Rome, or an English epitaph in a French 
 or German cemetery. 
 
 In addition to the numerals, the Etruscan epitaphs 
 will have furnished us with the following forms, which 
 I collect together here. If the interpretations were 
 undoubtedly true, the Etruscan question might pro- 
 bably be considered at an end. Where the Etruscan 
 forms are accompanied by a number, it is represented 
 by N. The proper name has to be supplied. 
 
 w " I'^tatisN." 
 avils JS ) 
 
 ril N, " annos N,f or " anno N." 
 
 line, "vivebat." 
 
 avil ril N, " anno aetatis N." 
 
 ril leine N) ,, . ., -x T „ 
 
 •7 A? 7 • c vivit annos JN . 
 ril JS leine ) 
 
 lupa, "obit." 
 
 lupu avils N ^ 
 
 avils N lupu > " obit aetatis N." 
 
 avils N livpuke ) 
 
 zilachnke, " sepelitur." 
 
 zilachnlce avils N, " sepelitur astatis N." 
 
 zilachnulce lupuhe, " sepelitur, obit." 
 
 avils machsNlupu, "obit aetatis mensis (mense) N." 
 
 avils thunesi N lupu, "obit aetatis diei (die) N." 
 
THE OLD ITALIANS. 71 
 
 1 . . avenke lupum amis machs N, " deponit cadaver 
 getatis monsis (niense) N." 
 eha suthi, " hie conditur." 
 eha suthi nesl, " hie conditur mortuus." 
 thui, " memoratur." 
 tular, "tumulus." 
 
 As a companion to the Etruscan epitaphs,, whose 
 
 examination I have now concluded, the longest of the 
 
 very few Phrygian epitaphs that we possess is here 
 
 subjoined, with its interpretation from the Armenian:* 
 
 Kelolces fenaftun aftas materes sosesait, 
 Celoces sepulcrum suse matris facit, 
 
 materes Efeteksetis Ofefinonomun. Lachit 
 matris Ephetexetis ex Ofefinone. J Sepelit 
 
 ( Vorat 
 
 ga mater an aresastin. Bonolc aJcenarvogafos 
 terra matrem prasstantem. Bonocus illustris 
 
 erekun telatos sostut; Inanon akenanogafos, 
 
 usum "Tsepulcri vetat ; Inanon illustris, 
 
 aer atanisen, Jcursaneson tanegertos. 
 vir judicialis, destructionem aedificii. 
 
 Cicero writes (De Leg., ii, 26) : cc De sepulcris nihil 
 
 est apud Solonem amplius, quam l ne quis ea deleat, 
 
 neve alienum inferat' ; poenaque est, ' si quis bustum 
 
 # It was said by Eudoxus (ap. Eustathiuin) : 'Ao/ueVioi to yivos 
 i>{ cfrpvyias, /cat t$\ (pavy iroXAa (ppvyi^ovdiv. As might be expected 
 from a contemporary of Xenophon and Agesilaus, and one who 
 was himself a traveller, Eudoxus is right about language : but the 
 Phrygians probably came from Armenia, not the Armenians from 
 Phrygia. 
 
72 THE ASIATIC AFFINITIES OF 
 
 (nam id puto appellari rvfifiov), aut monumentum, aut 
 columnam violarit, dejecerit, fregerit/" This passage, 
 as well as ancient epitaphs in general, will shew that 
 the Armenian brings out a good sense for the whole 
 of the Phrygian epitaph ; in which Bonok may have 
 been the lord of the ground, and Inanon a judge of 
 the district. Both Bonok and Inanon were illustres, 
 or " nobletnen," while Celoces was an untitled man, 
 perhaps a dependent or client of Bonok. 
 
 FenafUun, "sepulcrum," with an objective termina- 
 tion, is the Armenian anavth, or anoth, " vessel, pot, 
 box, area, ayyetov" and probably the Albanian ounth, 
 "a pot." 
 
 Aft-as, "suge," implies af, "he," which corresponds 
 to the Kurdish au, " this, that," and to the Armenian 
 iv, " he," which is found in ivr, " of him/' and also 
 "his." The genitive of ivr, "his," is ivroy. 
 
 The root of sosesait, "facit," may be discerned, 
 most likely, in the Armenian sos-aphel, "to handle, 
 manier" ; the composition of sosesait (and possibly its 
 root) in the Armenian sar-&s-el, " to form, to shape" 
 (root sar, in Persian sdz) ; and the conjugation of 
 sosesait in the Armenian t-ay, " he gives, dat" The 
 Phrygian, like the Latin, retains the final t, which is 
 dropped in Greek, Armenian, and Etruscan, as also in 
 Italian and Spanish. 
 
 Ofefinonoman, "ex Ofefinone," has a termination 
 which finds a parallel in the Armenian dmane, " from 
 this," and yaygma/i, " in the morning (ayg) " 
 
 Lachit is a verb of the i conjugation, which, if ren- 
 
THE OLD ITALIANS. 73 
 
 clered "vorat," is explained from the Armenian lake, 
 " he swallows "; and, if rendered " sepelit," contains 
 the root of the Etruscan zilachxike, which is found in 
 the Armenian etag, " fossa" [ante, p. 34). 
 
 Ga, " tellus/' is the Aryan word which appears in 
 Greek as <yr), in Gaelic as ce, in German as gait, in Sans- 
 krit as go, and in Armenian as Jcav, "clay/' and gav{ar), 
 "province.'" 
 
 Ares-astin, " praestantem," contains the Armenian 
 yarag, "before/' arag, "front/' arag-in, "first/' com- 
 bined with a termination like that of the Armenian nav- 
 ast, or n&v-asti, "a sailor." Indeed, ares-ast-in might be 
 obtained from the Armenian yarag -anal, "praBire, to 
 excel," just as ovr-ast, " a denier," is obtained from 
 ovr-anal, " to deny." 
 
 Erelc-un, "usum/' would contain the Greek ipy-, 
 and the Armenian erh, " toil/' herlc, " cultivation." 
 
 Aken-anogafos, "illustris," is derived from the same 
 word as the Armenian ahan-avor, "illustrious/' namely, 
 ahn, " an eye" ; and may perhaps contain also the 
 Armenian angov, "worthy of," so as to signify "worthy 
 of respect, honourable." 
 
 Telatos, if interpreted " sepulcri," may be allied to 
 the Armenian thai-el, "to bury," and thatar, "an 
 earthen vessel"; and to the Etruscan tular, "a tomb": 
 or, if interpreted "loci," to the Armenian teti, "place," 
 which seems akin to the root of thatar, namely, thai, 
 " tellus." The form of telatos is just like that of 
 Teparos. 
 
 Sostut, "vetat," is a verb of the ^conjugation, and 
 == Armenian saste, " he reprehends." 
 
 
 

 71 THE ASIATIC AFFINITIES OF 
 
 Aer (the word is doubtful) seems the Armenian 
 
 
 Atan-isen, "judicialis," contains the Armenian atean, 
 "tribunal, court of justice," and is equivalent in mean- 
 ing to the Armenian aten-alcan, "judicial, magistrate, 
 judge/' or to the Armenian aten-aJcal, "magistrate, 
 senator." 
 
 Kursan-eson, " destructionem," is explained from 
 the Armenian korzan-el, " to destroy," and korzan- 
 ovthivn, " destruction." The termination -es- rather 
 resembles the Armenian -ic u -tor," as in tov-ic, "giver, 
 da-tor." If kursaneson signifies "destroyer" or "in- 
 jurer," erehun ought to signify " user," not " use." 
 
 Tanegert-os, "sedificii," is a compound like Tigrano- 
 certa, and is composed of the Armenian words, tovn, 
 gen. tan, " a house," and kert, " building." 
 
 That akenanogafos is a title may be seen also from 
 the epitaph on the tomb of King Midas : — 
 
 Ates arkiaefas akenanogafos Midai gafagtaei fanaktei 
 edaes. 
 
 Here arkiaefas might signify "royal," from the Ar- 
 menian archay, "king"; or could perhaps be better 
 explained from the Armenian yarg, " value, esteem, 
 dignity," yargi, "respectable," yargoy, "honourable, 
 precious, reverend." Gafagtaei may be derived from 
 the Armenian gall, " throne/' and possibly be equiva- 
 lent to the Armenian gahakzi, the dative of gahakiz, 
 "sharer of a throne, fellow-sovereign" ; or else contain 
 the Armenian gta-nel, "to acquire, to have." For 
 fanaktei, "king," see infra, p. 79. Edaes, "posuit," 
 
THE OLD ITALIANS. 75 
 
 is a first aorist, like eOrjfce or e&ae, while the Arme- 
 nian edy "posuit," is a second aorist, like eOrj. The 
 terminations of the Armenian first aorist active, in the 
 third person singular, are -eaz and -az, according to 
 the conjugation. It usually wants the augment ; but 
 in e-k-eaz, "he lived/' and e-b-az, "he opened/' we 
 meet with forms completely like the Phrygian e-d-aes, 
 " posuit." Edaes terminates two other short inscrip- 
 tions, so that its meaning is clear. 
 
 I have already noticed the identity of the Phrygian 
 eXvfAos, " avkos" with the Armenian etegn, " reed." 
 Another Phrygian plant-name was remenia, " hyoscy- 
 amus, henbane/ 5 which is merely mentioned in my 
 Armenian dictionary as "a poisonous plant." But -eni 
 forms names of trees in Armenian, as in keraseni, " a 
 cherry-tree" ; and the same language has rem-akal 
 == okh-obkal, "malignant/' and therefore implicitly rem 
 == okh, "malice." Yet the nearest word to remenia is 
 the Sanskrit ramaniya, "pleasant, a charm," of which 
 the root is ram, " love, delight." Henbane may have 
 been used in philtres. An almost synonymous Sanskrit 
 word, priya. "gratus," might explain the first element 
 of the Dacian Trpta-hrfka, " ayu,7re\o? fiekaLva" of which 
 the second element, hrjXa, is the Armenian det, "herb, 
 medicine." The Phrygian crovaa, "Xzipia" is obviously 
 Semitic, but may have been derived immediately from 
 the Armenian sovsan, "lily." Zek/cta, the Phrygian for 
 Xayava, resembles more than one Armenian word : as 
 zatik (in composition zatk-), "flower"; zatk, "stalk"; 
 setkh, "melon," = Albanian salkyi. The relics of the 
 
76 THE ASTATIC AFFINITIES OF 
 
 Lydian language supply us with two names of the 
 same kind. One is clkvXov, " fiakavov Trpwivqv" which 
 may be compared with the German eicliel, and the 
 Armenian katin, "acorn." The other is fivcros, "6l;vr), 
 beech" ; a word which seems allied to the Abkhasian 
 mica, "wood"; to the Sanskrit micata, "sandal-wood"; 
 and to the Armenian mosay, " tamarisk." The name 
 of Mysia was supposed to be derived from juvcros. If 
 we may trust Horner, the tamarisk was common in the 
 Troad {II, vi. 39; x. 466; xxi. 18, 350). Herodotus 
 mentions it in Lydia (vii. 31), where Mr. Hamilton 
 (vol. ii, p. 144) speaks of "thickets of tamarisk." 
 There was a Mcesia Silva in Etruria, not far from 
 Rome. The Proper Thracian supplies us with three 
 plant-names. Bp/£a was the name of a plant like tlcj^tj, 
 of which the thema is rlfyos, "marsh, stagnant water." 
 The Sanskrit vrihi, and the Armenian brinz, "rice," 
 and the Rhaeto-Romance ritscha, " grass growing in 
 water," would be akin to j3pl£a. The second Thracian 
 plant-name is #77/409, " ocrirpiov n." The Sanskrit has 
 kdmin, " a climbing plant," and the Tuschi has ham, 
 = Georgian kama, " dill." The Bessi supply the third 
 Thracian plant-name, aaa, " firj^iov, tussilago, colts- 
 foot," which is plainly allied to the Armenian haz, 
 Sanskrit kdsa, "/3??f, tussis" ; and, as the Germans 
 call "coltsfoot" kuf -lattich, the Armenian hazar, "let- 
 tuce," might also be compared with haz and aaa. 
 The change of the Sanskrit Msa into the Armenian 
 haz is like the change, in Florentine pronunciation, of 
 casa into yasa or hasa. Professor Max Miiller has 
 
THE OLD ITALIANS. 77 
 
 noticed the apparent affinity of the Dacian fiavrela, 
 u sentis, rubies, bramble, blackberry/' to the Albanian; 
 in which language we find man, mande, " a mulberry, 
 a mulberry-tree/' mande pherre (pherre, " bramble"), 
 " a blackberry." In Armenian we have, according to 
 Rivola, mandak, iC genus herbae" ; and there is like- 
 wise the Armenian man-anekh, " mustard," and the 
 .Ossetic man-arth, "a raspberry, rubus ideeus." Per- 
 haps the original meaning of man-, mand-, or /jlclvt-, 
 is " berry" ; and the root may appear in the Sanskrit 
 mandala, " orbis, circuitus," and in the Armenian 
 man, "& round," which is also found in the expression 
 man pttoz, "blossom of fruits (pttoz) " All these 
 'words might have belonged to a race who spoke 
 dialects closely allied to the Sanskrit, as the Armenian 
 is, and who proceeded from a country in the position 
 of Armenia. The Dacian plant-names {ante, pp. 8, 9) 
 .have exhibited still more striking Armenian affinities. 
 
78 THE ASIATIC AFFINITIES OF 
 
 CHAPTER III. 
 
 ETRUSCAN VOTIVE INSCRIPTIONS. 
 
 The principle on which the argument in the preceding 
 chapter has been conducted is, to determine in general 
 the sense of each word without regard to its sound or 
 form, and then to see what language or languages can 
 explain that sense. This was possible in epitaphs, but 
 is not so in votive forms of expression, or at least not 
 so readily and completely. The shortest and, I think, 
 the best mode of proceeding here will be, to exhibit at 
 once the interpretations at which I have arrived, and 
 then to prove their correctness, or at least their fit- 
 ness.* I shall begin therefore by laying before the 
 reader the three longest Etruscan votive inscriptions, 
 accompanied by the interpretations which I put upon 
 them. The first is on a statue, now at Leyden (Micali, 
 Mon., tav. xliij; Lanzi, vol. ii, p. 455). It runs thus: — 
 
 Velias Fanahnal thuf-lthas alpan lenache 
 
 Veliae Fanacia-nats3 signum-precis supplex facessit 
 
 Men kecha tuthines tlen-acheis. 
 
 pia expiat gratiae debitum-pretium. 
 
 All the terms used here will be subsequently ex- 
 
 * The other or analytical process was adopted in my Armenian 
 Origin of the Etruscans (1861). 
 
THE OLD ITALIANS. 79 
 
 plained. The proper name, Fanah, deserves observa- 
 tion, on account of its resemblance to the Phrygian 
 fanahtei (ante, p. 74), == Greek clvcuc-tl, and of the 
 connexion of these words with the Armenian nahh, 
 "first/' and with the naqa, "king/' of the Persian 
 inscriptions. In the Milesian traditions reported by 
 Pausanias, Anax, = Armenian nahh, "first/' is the 
 Autochthon and the son of Earth (Ge). v Ava% would 
 be a Thracian, not a Hellenic word : at least it is 
 Armenian, and is not Latin. The father of St. Gregory, 
 the Apostle of Armenia, was said to be a Parthian 
 called Anah. 
 
 The next long inscription, on the base of the statue 
 of "the Orator'' in the Uffizi at Florence (Micali, Mon., 
 tav. xliv ; Lanzi, vol. ii, p. 468), is as follows : — 
 
 Aulesi Metelis Ve. Vesial Mensi hen 
 
 Aulus Metellus Veli filius Vesia-natus pius ut 
 
 fieres tehe sansl tenine tuthines chiseliks* 
 donum ponit libens fert gratis© monumentum. 
 
 The proper name, Vesia, may be compared with the 
 Armenian wes, " superbus." Ken, " ut/' seems found 
 in the Armenian hen, which is, however, only em- 
 ployed in composition. There are in Armenian two 
 triads of terms signifying "as, like": — 
 
 or-bar, or-pes, or-hen hi-bar, hi-pes, hi-hen. 
 
 Or signifies " who, which," and so would hi, which 
 the Armenian him, "quare," = Sanskrit him, "quare," 
 shews to be identical with the Sanskrit hi, "who, 
 
 # In the original, cMsvlilcs. I have changed the v (F) into e (E), 
 as in the case of tivrs, " thirty." 
 
80 THE ASIATIC AFFINITIES OF 
 
 which/' — Albanian Tcye. Bar means u modus" in 
 
 Armenian, and -pes = Sanskrit peg a y " forma." All 
 
 the six terms above therefore signify "quomodo"; and 
 
 ken may thus be interpreted " ut." 
 
 The following is the third long votive inscription, 
 
 and accompanies a statue of Apollo (Lanzi, vol. ii, p. 
 
 446 :— 
 
 mi fieres Epul { a )fo Aritimi 
 
 me donum Apollini et Artemidi 
 
 Fasti R(u)fr(u)a turlce Men kecha. 
 Fausta Rufria dat pia expiat. 
 
 Afe, " and/' would be identical with the Armenian 
 ev y " and/ 5 which Botticher refers to the Sanskrit 
 abhi, u ad, versus/' = Greek iirL But the word is 
 doubtful, and is not found elsewhere. Fabretti reads 
 svulare instead of Epul afe. 
 
 On a candelabrum (Lanzi, vol. ii. p. 421) there is 
 
 also this expression :■ — 
 
 Au. Velshus thup-lthas alp an turlce* 
 
 Aulus Veliscus signum-precis supplex dat. 
 
 I will next give a complete list of Etruscan votive 
 
 forms, omitting proper names, and reversing in three 
 
 instances, for the sake of comparison, the order in 
 
 which two words occur. The list would be this :— 
 
 * Aljpan turlce terminates two other votive forms. Fabretti, 
 1051, 1052. In the Etruscan kecha, "expiat," turk?, "dat," 
 suthi, "conditur," and lupu, "obit," we have examples of the 
 four Armenian conjugations in a, e, i, and u (ov); as in the 
 Armenian tay, "gives," at£, "hates," lmi, "is," and lizov, "licks." 
 Three of the four appear in the Phrygian sosesaii, lachi£, and 
 sostw£ (ante, pp. 72, 73), 
 
THE OLD ITALIANS. Si 
 
 1. kana- 
 
 2 . suthina 
 
 3. flezrl 
 
 4. fleres tlen-asies 
 
 5. fleres-zek sansl 
 
 6. turuke 
 
 7. fleres turke 
 
 8. fleres turJce klen kecha 
 
 9. alpan thup-lthas turke [acheis 
 
 10. alpan thuf-lthas lenache Men kecha tuthines tlen- 
 
 11. klensi fleres teke sansl tenine tuthines chise- 
 
 [liks 
 The next step in the argument will be to give the 
 meaning of these words, as deduced mainly from the 
 Armenian ; and to shew, as in the case of sepulchral 
 expressions, that such meanings make Etruscan votive 
 forms correspond in sense to the forms used by the 
 ancients. I therefore subjoin this comparative list of 
 ancient votive terms : — 
 
 LATIN & GREEK. ETRUSCAN. 
 
 'Rlfccov kana, " simulacrum, statua." 
 
 "AyaX/jia zek, " statua, figura." 
 
 "AvaOrfna suthina, cc sacrificiurn.-" 
 
 Eu%>}9 evetca thup-lthas, " signurn-precis." 
 
 Ex voto fleres, "donum." 
 
 Donum flezrl, " datum." 
 
 Votum 
 
 Libens sansl, "libens." 
 
 Men, klensi, "pius." 
 
 alpan, c( supplex, i/csTrjs, 9 ' 
 
82 THE ASIATIC AFFINITIES OF 
 
 LATIN & GREEK. ETRUSCAN. 
 
 dvedrj/ce tenine, "fer^" 
 
 Posuif teke> u ponit," or " facit." 
 
 Retulit 
 
 Fecit lenaclie, " facessit, fieri facit." 
 
 Declit turlce, " dat." 
 
 Solvit Icecha, " expiat," or " solvit." 
 
 Dedicavit 
 
 Consecravit 
 
 X.apt(TTrjpLa tuthines chiseliJcs, "gratis monumentum." 
 
 Merito tuthines tlen-acheis> "gratise' debitum- 
 
 [pretiuin." 
 tlen-asies, debitum- 
 
 [pretiurn.'" 
 
 The meanings assigned to tlie Etruscan words are 
 thus suitable, and therefore likely to be true ; and this 
 probability may be still further shown by the following 
 inscription, which is found in Grater, p. xlvii : — 
 Te precor* Alcide sacris invicte peractis 
 Ritef tuis Icetus dona ferens meritis% 
 Usee tibi nostra potest tenuis perferre caimna 
 Nam grates dignas§ tu potes efficere 
 Sume libens\\ simulacra^ tuis quse munera** cilo 
 Aris Urbanus dedicatff ipse sacris. 
 I shall now show how the meanings assigned to the 
 
 * Alpan f Klen, hlensi. 
 
 'I Sansl tenine tuthines chiselilcs, Tclen Icecha tuthines tlen-acheis, 
 5 Tlen-asies. || Sansl. % Kana, zeh. 
 
 ** Fleres, suthina ff Turke, Icecha. 
 
 
THE OLD ITALIANS. 83 
 
 Etruscan words are obtained, almost entirely from the 
 Armenian : — 
 
 Kana (1) " simulacrum" (Lanzi, vol. ii, pp. 465, 
 466). Gaelic caon, " simulacrum ;" or Armenian Jc-al, 
 " sistere," -an, Armenian termination, = Sanskrit 
 -ana, Kana, is found on statues. 
 
 Suthina (2), (C sacrificium." 
 
 Tuthines (10, 11), " gratiae, donationis, yapiTos" 
 These words are of the utmost significance. Suthina 
 is found alone on a number of objects ; among others, 
 on a statue (Micali, Mon., tav. xxxy, 9), and on the 
 back of a patera (tav. xlviii) : it is also sometimes ac- 
 companied by a proper name (Fabretti, p. clxxxiii). 
 The most probable meaning of suthina is obviously " a 
 votive offering;" and it would therefore be duly ex- 
 plained from the Sanskrit hu, "Diis offerre, sacrificare/' 
 just as the Etruscan suth-i, " is buried," was explained 
 from the Sanskrit hud, " coacervare, submergi" (ante, 
 p. 62). Though it would seem probable, at first 
 sight, that suthi and suthina were cognate words, yet 
 they need be no more so than potis and potio ; lot, lotus, 
 and lotion ; rat and ration. We may also fairly con- 
 clude that suthina does not mean " tomb" or " urn ;" 
 for such an inscription as Larth Seties suthina is found 
 on several vases, " in nonnullis vasis" (Fabretti, 2095 
 quinq. B). The ashes of a deceased person would not 
 be distributed among a number of vases : nor, indeed, 
 does there seem any sense but " votive offering" which 
 will explain the fact of suthina being found as a single 
 word on statues, paterae, and, as in Fabretti, quinq. A, 
 
8 1 THE ASIATIC AFFINITIES OF 
 
 " in nonnullis monumentis aheneis," discovered at one 
 place. As therefore sathi appears on tombs, it is pro- 
 bable that suthi and suthina have no etymological con- 
 nexion ; and it is also probable that, while suthina is a 
 nominative, tuthines is the genitive of a similar nomina- 
 tive tuthina, where the termination is the same as that 
 of suthina, but the root different, just as we find in 
 Latin natio and ratio ,motio and notio. We should thus 
 have these words in Etruscan : — 
 
 su-thina, or else s(u)-uthina, " sacrificium." 
 tu-thina, or else t(u)-uthina, cc ....... " 
 
 In the termination -uthina, an Armenian would at 
 once recognise his native termination -ovthivn, which 
 is so common, as to occur three times in the Lord's 
 - -.Prayer, as well as in twelve nouns derived from nav, 
 /'"a ship/' and in fifteen derived from mard, "a man.-" 
 Th^ Sanskrit hu, "Diis offerre, sacrificare/' becomes the 
 Armenian zoh, which would be suh in Etruscan ortho- 
 graphy • while the root of the Etruscan t(u) -uthina or 
 t-nthina is found in the Armenian tov or t-, " give ;" 
 for u d-are" is t-al, and " d-atus" is tov-eal, in Ar- 
 menian. We have, indeed, both the Etruscan words 
 in Armenian : for " human sacrifice" is marda-^o/iov- 
 thivn, and " giving of homage" is haxka>-tov ovthivn, in 
 that language. 
 
 The final vowel in suthin-a would be dropped in Ar- 
 menian, as may be seen by such an instance as kin, 
 " yvvij" with which may be compared the Etruscan 
 Una, that occurs in the beginning of an inscription at 
 Volterra : — 
 
THE OLD ITALIANS. 85 
 
 Titesi Kalesi Mna Ks Mestles 
 
 The Etruscan proper name ; Mestles, resembles the 
 name of the Maeonian leader, M.ea6\7}s (II. n, 864), and 
 that of the Iberian town, Meo-rXrjra, mentioned by V 
 Ptolemy. Mes- may be the Armenian mez, " great," 
 — Zend mazo, = Sanskrit mah(at), another example 
 of the change of the Sanskrit h into a sibilant. 
 
 The following table will exhibit the affinity between 
 the Armenian and Etruscan in several points already 
 considered ; and may likewise explain, to some extent, 
 by showing in what manner the Armenian uses infini- 
 tives and participles as nouns, how the Etruscan comes 
 to have so many I terminations : — 
 
 AEMENIAN EOOTS & WOEDS. ETEUSCAN EOOTS & WOEDS. 
 
 sovz, " cond-ere." suth, = Sanskrit hud. 
 
 zoh, " sacrifice." su, = Sanskrit hu. 
 
 tovy "give." tu. 
 
 av, "age, increase." av, = Sanskrit av. 
 
 rah, " go." ri, = Sanskrit ri. 
 
 tes, " sight." 
 
 tes-ovc, gen. tes-ci, "inspector." 
 
 tes-avor, " apparent." 
 
 tes-il (videri), "appearance." av-i7, "age." 
 
 tes-avor-il, "to appear." r(i)-iZ, "year." 
 
 tes-an-el, " to see." 
 
 tes-an-el, "appearance." 
 
 tes-ot (videns), "prophet." 
 
 tes-an-ot (videns), "prophet." 
 
 tes-ovthivn, " sight." s (u) -uthina, "saerificium." 
 
 tes-c-ovthivn, "inspection." t{a) -uthines, "donationis." 
 
80 THE ASIATIC AFFINITIES OF 
 
 tes-an-el-ovtliivn, "visibility." 
 tcs-ot-ovtliivn, " sight." 
 tes-an-ot-ovthivn, "sight." 
 tes-an-eli, " visurus, videndus." 
 tesanelich, "the eyes, sight." 
 
 These last two Armenian forms will illustrate in its 
 place the Etruscan chiseliJcs. 
 Flezrl (3), "datum." 
 Meres (4, 5, 7, 8, 1,1), "donum." 
 
 I have met with flezrl only once : it occurs on the 
 
 back of a statue (Micali, Mon., tav. xxxiii). The 
 
 analogy of fleres would lead us to expect flerzl instead 
 
 of flezrl, while we should infer from nesl, " mortuus," 
 
 that flezrl or flerzl is probably a participle. On a patera 
 
 in Lanzi (tav. xi), a pedestal with a bust is inscribed 
 
 flere. The connexion with^eo and ploro is most likely 
 
 a correct one. The Armenian gives :- — eter, " fletus" 
 
 — aters, "precis" — ovterz, " donum" — and eterzeal, 
 
 otorzeal, and ovterzeal, " datus, oblatus." The initial 
 
 vowels here may be due to the circumstance, that few 
 
 Armenian words are allowed to begin with t, = ^\, = 
 
 Welsh 11, which the English pronunciation converts 
 
 into till or fl. The Armenian etag {ante, p. 34) is an 
 
 instance of a vowel being prefixed to t. The/ in fleres 
 
 may represent the aspirate % contained in t. But 
 
 compare also the Armenian lov, "flea, floh," where / 
 
 before I is entirely dropped. In li, " 7rA,e-o?," p is 
 
 dropped. 
 
 Tlen-asies (4) ) ^ i i -x l* » 
 
 Tlen-acheis (10) \ debltum P retmm - 
 
 Gaelic dligh, " debe ;" dlighe, "lex, debitum ;" dleas, 
 
THE OLD ITALIANS. 87 
 
 " officium ;" root dl-, in Etruscan tU : -ean, Armenian 
 
 adjectival termination. 
 
 Armenian azech, "pretium," a plural noun: in the 
 
 objective, the final -ch becomes -s. Ossetic achos " a 
 
 sum due ;" achza, " money." Greek a%ia. Compare 
 
 ac/ieis and asies with the two Armenian forms for 
 
 " daughter," doMt and dovstr. 
 
 Turuke (6) } (( -, , ^ „„ 
 
 Turke (7,8,9) / dat - S ^«- 
 
 Armenian tovrch, " gift," a plural noun, of which the 
 root or base is tovr : it is found in phar-a-tre, " he 
 gives glory (phar)" As already noticed, a great many 
 Etruscan verbs seem to terminate in -Ice, like tur-ke ; 
 a form apparently used by the Lydians, as the Lydian 
 fiao-Ke, " e%€06a%€," would = the Armenian waz-er, "he 
 was rushing." Another example of such a form is 
 supplied by the picture which forms the frontispiece to 
 the second volume of Mr. Dennis' Etruria. It repre- 
 sents the self-devotion of Alcestis to death for her 
 husband, and is accompanied by this inscription ; — 
 eka erske nak achrum flerthrke. 
 
 Eka has already been interpreted "here" or " ecce" 
 from the Armenian aha, " lo !" The Georgian has aha, 
 "ecce;" ach, "hie;" and acha, "ibi." Ers-ke, as we 
 know the subject of the picture, may be considered as 
 equivalent to the Armenian eres-e, " she offers herself." 
 Nak may mean " to," like the German nach, the Hun- 
 garian nak, and the Tuschi naqw ; and a similar word 
 would appear in the Armenian nakh, "first, before," 
 which is both adjective and adverb, and has in com- 
 position nearly the force of the German nach, as in the 
 
 
 yt* 
 
88 THE ASIATIC AFFINITIES OF 
 
 Armenian nakhanz, " envy." And thus, as fler-thrlce 
 would be already explained under fleres (p. 86) and 
 tiirke (p. 87), the whole inscription would probably 
 mean : " Lo ! she offers herself to Acheron as a de- 
 voted (or c suppliant') gift." As a comparison of forms, 
 notice that thrhe elides the vowel in tiirke, as the Ar- 
 menian srbe, " he sanctifies/' elides the vowel ov, = u> 
 in sovrh, "holy," = Sanscrit guhlira, " splendidus, 
 albus." Compare the Sabine cyprum, "bonum," and 
 the Etruscan Cypra, " Juno/' which would be a Sabine 
 word. 
 
 Zek (5), u statua, figura." 
 
 Armenian zev, zevak, "form, figure." Zek might 
 likewise be rendered " brought," from the Armenian 
 zcjely = German Ziehen ; or " produced/' from the 
 Armenian zagel, = German zeugen.. Perhaps zek, 
 which only occurs once, should be zeke, " brings," = 
 Armenian zge. 
 
 Sansl (5, 11), "libens." 
 
 Armenian znzavt, znzot, or znzot, "gaudens, 
 libens." The inscription (5), when completed, is : fleres 
 zeh(e) sansl kver. It is found on the statue of a boy. 
 I should interpret kver, "soror," guided by such Latin 
 inscriptions as : — 
 
 D.M.C. Egnatio Epicteto et C. Egnatio Floro 
 modesta soror. 
 
 Fortunato fratri pientissimo fecerunt sorores. 
 
 The Armenian for " sister" is choyr, which is pro- 
 nounced like the English queer, and gives cher in the 
 genitive. The Persian is khwdher ; the Welsh, chwaer ; 
 
THE OLD ITALIANS. 89 
 
 and the Breton, choar. On another statue of a boy, 
 with the inscribed arm unfortunately nearly broken off 
 (Micali, Mon., tav. xliv), is this fragment of an inscrip- 
 tion :— 
 
 as velusa 
 
 is kelvansl 
 
 ..... s lever thvethli 
 
 klan 
 
 As the Etruscan Thanchvil is the Latin Tanaquil, the 
 Etruscan lever would probably be represented by quer 
 in Latin. In the votive inscription, fleres tlen-asies 
 sver, sver may = hver, as tlen-asies seems = tlen-ac/ieis 
 (ante, p. 86). 
 
 If the Etruscan sansl and nesl be "libens" and 
 ' c necatus," and if avil and ril be infinitives as well as 
 nouns, we should find these terminations in four 
 languages : — 
 
 ETRUSCAN. ARMENIAN. PERSIAN. SANSKRIT. 
 
 Infinitive ... -I -I -dan -turn 
 
 Pres. Part.... -I -t -n -fnjt 
 
 Past. Part.... -I -I -dah -ta 
 
 The Lydian fcav$av\(7]<;) would exhibit the present par- 
 ticiple in a third Thracian dialect (ante, p. 7). 
 
 Klen (8, 10) \ " pius, rite," i.e. " with due religious 
 Klensi (11) j rites." 
 
 For the termination of Mensi, which distinguishes it 
 
 from Men, compare the Armenian layn and lay mi, 
 
 "broad ;" or bolor, " a circle," and bolorsi, "round;" 
 
 and for the meaning of Men, compare the Gaelic glan 
 
 and the Welsh glan, cjlain, "pure, holy, clean, beau- 
 
90 THE ASIATIC AFFINITIES OF 
 
 tiful, fair ;" and perhaps the Armenian getani, " fair, 
 
 decent." 
 
 On a tomb at Vulci (Micali, Mon. Ined., tav. lix) the 
 
 sculptured figure of a man stands in a rock-hewn blank 
 
 doorway, and is surrounded by an inscription which 
 
 may be interpreted as follows : — 
 
 Eka suthik Veins Ezpus klensi kerinu 
 " Hie tumulum Velus Bzpus pius sculpit." 
 
 For suthik may be considered as a diminutive of 
 suthi, cc a tomb" (ante, p. 69), like the Armenian 
 masnik, " a particle/' from masn, cc a part ;" or lovsik, 
 " a, little light," from loys, " light." Ker-in-u, like 
 lup-u {ante, p. 33), would be an Armenian verb of 
 the ov or u conjugation, derived from a root ker, which 
 may be allied to cher-el, " to scrape," or ger-el, " to 
 write," or k'er-el, " to hammer, to carve." For the -in* 
 in kerinu, compare Armenian forms like lizane, lize, 
 lizov, lezov, "he licks" — gotanay, gote, "he steals"— 
 kheranay, kheri, "he insults" — kamenay, kami, " he 
 wishes." Similar n forms are common in Aryan lan- 
 guages. Lini, " he is," supplies another instance in 
 Armenian. 
 
 Kecha (8, 10), "expiat, consecrat," or else "solvit." 
 
 Armenian chahe, " expiat" — chake, " solvit." 
 
 There is also kahe, "parat." See ante, p. 37, No. 2340 : 
 
 Malum he . . . , " funera " 
 
 Tenine (11), "fert, offert." 
 
 Armenian tani, "fert, reddit, tenet" Ten-in-e would 
 be a form like ker-in-u, just noticed, but in an e, not a 
 u conjugation. The simpler form is tenu (ante, p. 36, 
 
THE OLD ITALIANS. 91 
 
 note f). Compare tlie Armenian psnov, = psne, = pse, 
 "he considers/'' 
 
 Alpan (9, 10), "supplex, l/cerrj^, flehend." 
 
 Armenian otb, " fletus :" -an, -ean, Armenian ad- 
 jectival terminations. The letters o and b are wanting 
 in Etruscan. 
 
 Tele (11), "facit," or else "ponit." 
 
 Lapponic iakk-et, Fin tek-'d, " facere/' Sanskrit tales, 
 u facere, fabricare, findere f taksan, cc faber lignarius." 
 Armenian thak, " a hammer, a mallet •" thak-el, " to 
 beat, to ram/' Latin tignum, tigillum. Greek re/crccp, 
 
 Te-ke might also be regarded as an Etruscan verb in 
 
 -Tee from a root t- " placing." This Aryan root is in 
 
 Armenian d-, and de-ne is " ponit," an n form like the 
 
 Etruscan ten-ine and Jcer-inu. 
 
 Thup-lthas (9) ) cc . . , ., 
 
 Thuf-Uhas (10) | S1 « num P recis > an ex vot0 - 
 
 Tv7ro9 Xm)?. Armenian tophel, thopel, dopliel, 
 
 " Tvirreiv" -atothel, "precari;" itz, c ' desiderium ;" 
 
 etzal, " desiderare." Though the Armenian avoids t 
 as an initial, yet we find tzali, as well as etzali, " de- 
 siderandus."" Two? illustrates teke by exhibiting the 
 connexion between ce striking," and " forming" or 
 cc making." As tu7T09 and Xlttj are not Latin (for 
 typus is borrowed), they would be Thracian rather than 
 Hellenic words, if the Hellenes were an offshoot of the 
 same Italian race to which the Umbrians and Oscans 
 belonged. 
 
 Lenache (10), "facessit." 
 
92 THE ASIATIC AFFINITIES OF 
 
 Armenian etanake, " modulatur ;" etanak, " modus, 
 forma ;" etanil " fieri ;" linel, " esse,, fieri, existere." 
 There is, besides, the suffix -eten, "fac(tus)," as in 
 osketen, " made of gold fosJci)".* In Armenian, the 
 termination - ak is frequently causative, like the Sans- 
 krit -oka: e.g. e 9 " existence ;" eak, " creator;" so 
 that, as lin-el is " fieri," lin-ake would be " fieri facit." 
 If lenache be " facessit," teJce would be " ponit" rather 
 than "faeit." 
 
 Ghiseliks (11), " monumentum, in memoriam." 
 
 Armenian yiselich, " a memorial," the plural form of 
 yiseli, of which the diminutive would be yiselik (a form 
 like the Etruscan suthik), and its plural form yiselikch, 
 in the objective yiseliks. Yiseli is the future participle 
 of yisel, "to remember," of which the root is yis. 
 Similar forms to yiselich are :—talich, "gift," Iselich, 
 "ear, audience," empelich, "beverage," and tesanelich, 
 "sight, eyes" {ante, p. 86). But the existing Arme- 
 nian forms derived from the root khat, " playing," will 
 most clearly exemplify the supposed formation of 
 chiseliks from a root chis. 
 
 ARMENIAN. ETRUSCAN. 
 
 khat, "ludus." chis 
 
 khatal, " ludere." 
 
 khat ali, "ludendus." 
 
 khaialich, "ludus," i.e. "ludenda." 
 
 khatalik, " ludus" (dimin.) 
 
 khataliks, "ludos." chiseliks 
 
 For the affinity between the Etruscan chis and the 
 
 Armenian yis, compare the Armenian khovzel, yovzel, 
 
 # Cf. Lithuanian auksas, Prussian ausis. 
 
THE OLD ITALIANS. 93 
 
 "to seek." The Armenian y is aspirated.* In the 
 mountains of Noricum there was a place called Gand- 
 alicce, which resembles the Armenian khat-alik in form, 
 and which may, by the aid of the root eand, "throttle" 
 (ante,ip. 7), be interpreted "gorges/ 5 or " etrangle- 
 ments." The Albanian has erseli, "honourable" (th. ers, 
 "honour"), to compare with the Armenian yargeli, 
 ."venerandus." On the Noric Elegium, see ante, p. 66. 
 
 I have now completely gone through the votive words 
 comprised in the forms at the beginning of the chapter 
 (ante, p. 81). A few other words may be added to 
 them. The first is the well-known Tins-kvil, which 
 stands alone on three votive offerings (one of them the 
 celebrated Chimaera), and in which the name of Tina, 
 the Etruscan Jupiter, has long been recognised. Tina 
 would be the Sanskrit dina, "day/' a contraction of 
 divana (Lassen) = divan, " day." The root is div, "to 
 shine," which appears in the Armenian tiv, " day," in 
 the Latin dies, divum, divinus, and, as already noticed, 
 in the Albanian diet or dil, "the sun." It may be re- 
 marked here in passing, how the Albanian velazer, 
 " brother," (which the plural shows to be the complete 
 form of vela, " brother"), enables us to pass from the 
 Sanskrit bhrdtri to the Armenian etbayr, for btayr. 
 
 The Etruscan hvil in Tins-kvil seems = Armenian 
 khilay, "gift;" which would make Tinskvil signify 
 " Jovis donum, Jupiter's gift, a gift to Jupiter/ 5 
 
 This word is found, together with some others, on 
 
 * We have also the Etruscan kiem, chiem, " five/' to put by the 
 side of the Armenian hing, "five/' and yi-sovn, "fif-ty." 
 
94 THE ASIATIC AFFINITIES OF 
 
 the beautiful candelabrum of Cortona. The inscription, 
 which has slightly suffered from a fracture, appears to 
 have run thus : — 
 
 thapna lusni (T)inshvil Atlili(i) sulthn 
 
 As Ath and Athl are both proper names in Etruscan 
 
 (Lanzi, torn, n, p. 363 : — there is an Armenian district 
 
 called Athli), the meaning of the inscription may be — 
 
 u A burner of light, offered to Tina, the work 
 
 of Atilius." 
 
 Thapna, " tcavarris"— lusni, "lumims," — and salthn, 
 
 " cast, fonte, formatio, opus/' may be thus explained : 
 
 1. Sanskrit tap, "burn," = Armenian tap, thaph : 
 Sanskrit svapm, " sleep ;" Armenian masn, " part/-' 
 otn, " foot." 
 
 2. Armenian loys, c c light •" lovsin, gen. lovsni, "the 
 moon / ; * lovsn-thag ("light-crown"), "the planet 
 Jupiter." 
 
 3. Salthn might be explained either from the Ar- 
 menian sat-mn, " an embryo," or from the Armenian 
 sat-el, " to mix, to knead." The termination -thn may 
 be a contraction of -ovthivn (ante, p. 84). In form, 
 salthn resembles the Armenian sovrthn, " orifice." f 
 
 1 On a patera or mirror in Lanzi (tav. xn, No. 6) Diana is 
 called Losna, sl name remarkable for containing the non-Etruscan 
 vowel o. 
 
 f The Etruscan kech- (ante, p. 90) and the Armenian chali- (or 
 chav- ; cf. Sanskrit Jihav, " purificari") would explain two words in 
 Hesychius, of which the last might exhibit the I termination of 
 the Armenian, Lydian, and Etruscan present participle (ante, p. 
 89):— 
 
 Kqi(k]s) or k6(t]s), " Upevs Kafieipcav 6 naQalpwv cpovea" 
 Koi6\(r}s), " fepetfs." 
 
THE OLD ITALIANS. 95 
 
 CHAPTER IV. 
 
 THE INSCRIPTION OF CERVETKI. 
 
 In the two previous chapters, which have been devoted 
 to the consideration of sepulchral and votive forms in 
 Etruscan, .the force of the argument is in a great 
 measure derived from the fact, that the Armenian 
 language enables us to explain the Etruscan words in 
 such a manner as to make the sense of the Etruscan 
 forms correspond closely to that of other ancient forms 
 of the same kinds. The meaning assigned to the 
 Etruscan words may sometimes be described as certain, 
 as in the cases of avil, ril, and leine, and may be 
 generally affirmed as more or less probable in every 
 case ; so that the argument in favour of the Armenian 
 or Thracian affinities of the Etruscan becomes very 
 strong. In the subject of the present chapter, there 
 are no such analogies to guide us : there is no sense 
 which we are bound to elicit from the Etruscan by the 
 aid of the Armenian, if the intimate relationship be- 
 tween the two languages is to be maintained. That 
 the inscription of Cervetri is Armenian, depends chiefly 
 upon the singular closeness with which the Armenian 
 fits it, and which is such that even the metre of the 
 inscription, for it is written in verse, scarcely suffers at 
 
96 THE ASIATIC AFFINITIES OF 
 
 all, while a good and appropriate sense is brought out 
 for it at the same time. 
 
 The Inscription of Cervetri, the ancient Agylla or 
 G cere j is written on a pot or cup of antique black 
 ware, capable of holding nearly a pint ; and it con- 
 sists of two hexameter verses, but with the words all 
 run together. They should probably be divided thus : — 
 
 mi ni hethu ma mi mathu mar am lisiai thipurenai 
 
 ethe erai sie epana mi nethu nastav helephu. 
 
 As it might be objected that the lines are divided 
 into words so as to adapt them to the Armenian, it 
 may be as well to mention that my division of them 
 is the same as that of Lepsius, with the exception that 
 he reads mar am instead of mar am in the first line, 
 and minethu instead of mi nethu in the second, where 
 I follow Dr. Donaldson. Maram may perhaps be 
 preferable to mar am, as will ultimately appear : but, 
 if we read mi ni hethu and mi mathu with Lepsius, 
 then mi nethu seems more probable than minethu. As 
 we have already met with, mi, " I," in Etruscan, it 
 may reasonably be conjectured that the cup is made to 
 speak of itself, and that it affirms of mathu what it 
 denies of kethu. Again, as the Etruscan is an Aryan 
 language, it will be at once suspected that mathu 
 means "wine," for such a word occurs in a great 
 number of Aryan dialects, from the English mead to 
 the Sanskrit madhu. 
 
 With this slight clue to the tenor of the inscription, 
 I will now proceed to interpret it word for word, as I 
 have divided it : — 
 
THE OLD ITALIANS. 97 
 
 1. Mi, "I." 
 
 The Armenian for "I" is es ; for a me/' (z)is : but 
 me, " I/' and me, " me/' exist implicitly in the Arme- 
 nian plurals, mech, "we" (= Lithuanian mes), and 
 [z)mez, "us." For dov, "thou/' makes dovch, "ye"; 
 and the Armenian nominative plural is formed by the 
 addition of -c/i to the singular, and the accusative 
 plural by that of -s. 
 
 Welsh and Gaelic mi, "I"; Georgian me, "I": etc. 
 
 2. Ni, "not." 
 
 Armenian mi = Greek //^ = Latin we. 
 Welsh and Gaelic m, "not." Behistun Persian 
 niya, " not." Persian mail, noli, " not." 
 
 3. Eethu, "of water"; less probably, "of milk." 
 The Armenian get, " river," hath, " drop," kith, 
 
 " milking/ 5 and kathn, " milk," are thus declined : — 
 
 Nominative get* hath Mth kathn. 
 
 Genitive getoy kathi kthoy kathin. 
 
 Dative ( J e ^°y kathi kthoy kathin. 
 
 Ablative g e toy kathe kthoy kathine. 
 
 Instrumental ...getow kathiv kthow kathamb. 
 
 There is in Etruscan prose a remarkable deficiency 
 of vowels, which does not appear in the Inscription of 
 Cervetri ; a difference which has led to the inference 
 that the language of the Inscription of Cervetri is not 
 Etruscan, f But this inference is too hasty, for the 
 
 * Armenian infinitives are declined completely, like get. 
 
 f " This (the Inscription of Cervetri) is neither Latin, nor Greek, 
 nor Umbrian, nor Oscan. It is equally certain that it is not 
 Etruscan ; since in that tongue harsh unions of consonants 
 
 H 
 
98 THE ASIATIC AFFINITIES OF 
 
 same peculiarity exists in Armenian, where the vowel 
 Sj which has the same sound as the Sanskrit a, is 
 continually understood in prose, where it would be 
 expressed in poetry. Aucher says of this Armenian 
 letter : " Entre deux ou plusieurs consonnes elle est 
 to uj ours sousentendue ; mais dans la division des 
 mots elle se met actueUement, comme aussi dans la 
 poesie" Sir Henry Rawlinson, in his explanation of 
 the Persian Cuneiform Inscriptions (p. 55), writes : 
 " The short sound of a was optionally inherent in all 
 the consonants of the (ancient) Persian alphabet. This 
 principle of organisation is common to every single 
 branch of Arian Palaeography, with the exception of 
 the Zend." The Armenians and Etruscans exercised 
 such an option by dropping in prose the short a, 
 = Armenian e, while the Phrygians preferred ex- 
 pressing it. In Sanskrit, the short a is necessarily 
 implied in every consonant or combination of con- 
 sonants, when unaccompanied by the mark Virdma, or 
 the sign of some other vowel. 
 
 An Etruscan word where the deficiency of vowels is 
 particularly great is trutnvt, which seems in a bi- 
 lingual inscription (Lanzi, vol. ii. p. 565) to correspond 
 to haruspex, and may be composed of the root or base 
 
 abound, while in this the distribution of vowels is as well propor- 
 tioned as in the Negro-languages : moreover none of the well- 
 known Etruscan words here occur." — Newman's Regal Rome, p. 7. 
 Mi, " I, me," does occur in the Inscription of Cervetri : nor is there 
 any reason why two verses on a drinking-cup should contain any 
 votive or sepulchral words. 
 
THE OLD ITALIANS. 99 
 
 of trutinor, with the Armenian termination -ovaz, 
 — Sanskrit and Zend -vat. But though in trutnvt there 
 is only one vowel to six consonants, yet the same pro- 
 portion is observed in two Armenian words of similar 
 sound: in trtngel, "to murmur/' and in 6r9nguk, *^ A ^- 
 " sorrel/' gen. Ordnghi.* In two of the Armenian 
 words cited above to explain the Etruscan ketha, namely 
 Jcthoy and Mhow, there is an e deficient, and they 
 would be written in poetry Jcethoy and Jcethow ; or in 
 Etruscan orthography, as the Etruscans had no o, 
 hethu in both cases. Getoy and getow would become 
 in like manner Jcetu, as the Etruscans had no medial 
 consonants. 
 
 Gaelic eith, gith, " imber" ; ce, gen. ceithe, " flos 
 lactis." Albanian cheth, "stillare, fundere." Latin 
 gutta. 
 
 4. Ma, "but?' 
 
 Armenian na, "but, however, rather, in fact, verum" 
 Sanskrit dm, "nee" Tuschi ma, "but." Lapponic 
 ma, " quidem." 
 
 5. Mi, " I." 
 
 6. Mathu, " of wine." 
 
 Armenian math," syrup of grapes, raisine, defrutum" 
 which is declined like get (3). Mathoy would become 
 mathu in Etruscan. 
 
 German meth : English mead : Welsh medd : Greek 
 
 # I take this word to be the TepTavdyera of Dioscorides, one of 
 the five names by which the Eomans knew artemisia or mugwori. 
 Like Pliny's elegia {ante, p. 66), it might possibly have been bor- 
 rowed from Etruria. 
 
TOO THE ASTATIC AFFINITIES OF 
 
 fjiedv : Zend mathu, "wine": Sanskrit madliu, "honey, 
 wine, intoxicating drink"; mad, "to be intoxicated" ; 
 mada, "intoxication, madness": Persian may, mul, 
 " wine" : Lydian fio)Xa^, " elSo? oivov." Armenian 
 m.oli, "mad, intoxicated"; metr, "honey/' gen. metov : 
 Greek /ulwXv. 
 
 7. Mar, "a pot" or " measure" (German mass). 
 Armenian mar, "a measure of liquids" — " fierprjTrjs, 
 firkin" (John ii. 6): Persian mar, "measure, number": 
 Greek fidpts, "a measure containing six /corvXac (about 
 three pints)":* Albanian mere, "every liquid and dry 
 measure": Lithuanian mera, "measure": Russian mjera, 
 " measure." Albanian marr, " to hold, to contain": 
 Georgian marani, "a wine-cellar," = Armenian maran, 
 Sanskrit, met, mas, "to measure." The Etruscan mar 
 and mathu seem to contain Aryan roots of universal 
 prevalence. 
 
 8. Am, " 
 
 am." 
 
 
 Armenian. 
 
 ...em 
 
 
 Persian . . . 
 
 ...am 
 
 
 Behistun ... 
 Zend 
 
 ...amiya 
 ...ahmi 
 
 > "lam." 
 
 Albanian . . . 
 
 ...yam 
 
 
 Greek 
 
 Sanskrit ... 
 
 f r 
 ...€iflL 
 
 . . . asmi 
 
 
 9. Lisiai, 
 
 <* for the t 
 
 3ngue." 
 
 Armenian 
 
 . . . lezov 
 
 ) 
 
 Lithuanian 
 Hebrew . . . 
 
 . . . lezivwis 
 ...lason 
 
 > " a tongue. 
 
 * Mdpis may be a Thracian word ; and the Latin dolium appears 
 in like manner = Armenian doyl, if bucket." 
 
THE OLD ITALIANS. 101 
 
 ( lezovl 
 
 A . ) lizovl 
 
 Armenian ...< 7 . 7 
 J lizeL 
 
 \ lizanel 
 
 Lithuanian lezu 
 
 Persian lisidan 
 
 Persian ...lis 
 
 Sanskrit lih 
 
 y " to lick. 
 
 lick" (root). 
 
 Observe the letter-change in the Sanskrit lih and 
 the Etruscan 7?'s-iai. The genitive and dative of the 
 Armenian lezov (i. e. lezu) are lezovi. But the Etruscan 
 lisia, " a tongue/' would be declined more nearly like 
 such borrowed Armenian proper names as Angtia, 
 ''England/' gen. and dat. Angtiay; or Ananla, "Ana- 
 nias/' gen. and dat. Ananiay. 
 
 10. Thipurenai, "for the thirsty" (tongue). 
 
 In thip we meet with a very common Aryan root 
 for " heat." In Armenian this root is tap or thaph, 
 which has been discerned before in the Etruscan 
 thapna (ante, p. 94). The Armenian tapean, "burn- 
 ing, heated/ 5 would give the meaning of thipurenai, 
 but the termination must be explained from such 
 Armenian words as those which follow : — 
 
 ( hayr, " father." _ Y „ 
 
 \ hayr -or en, " paternally." 
 
 J archay, " king." 
 
 ( archay-oren, " royally." 
 
 ( ham-ah-, "entire, entirely": root ham, = 6/^(09). 
 < ham- or en, " entire, entirely." 
 ( ham-orini, gen. and dat. of ham-oren. 
 
 J get, " beauty." 
 
 \ get-a-yoren, " fair." 
 
102 THE ASIATIC AFFINITIES OF 
 
 ren, u a law, a rule." 
 drin-ali, "example, type, form." 
 yorin-cl, " to form, to shape." 
 orin-akel, "to form, to represent."* 
 
 Nearly similar terminations may be found in the 
 Armenian words : — Hay-even, " Armenian, Haican" — 
 pliokli-ar£n, " payment " — kerp, herp-aran, " form, 
 figure." There is no indication of genders in the 
 Armenian language; but such proper names as Athenas, 
 " Pallas," gen. and dat. Athenay, are declined like 
 thipurenai, which would be a feminine adjective agree- 
 ing 1 with lisiai. 
 
 11. mite, "if," or "when." f 
 
 Armenian ...etlie 
 
 Zend yezi, yeidhi I <c . ~ ., 
 
 Behistun yadiya 
 
 Sanskrit yadi 
 
 * Compare with. 6rin~aM, " he forms/' the Etruscan verbs, len- 
 ache, tur-Jce and iur-uke, ers-ke, te-ke, zilachn-ke and zilachn-uke. 
 Such Armenian words as hayr, " father/' and mayr, " mother/' in- 
 timate that the Armenian language, in its earliest existent state, 
 is not very ancient. Indeed, the oldest Armenian writings only 
 date from about 400 a d. For the Armenian forms for " daughter/' 
 " brother/' and " sister," see ante, pp. 63, 88, 93, The Phrygian in- 
 scriptions, as might be expected, bear evidence of much higher 
 antiquity ; for in them we find materes and materan, — fiyrepos and 
 [jL-riTdpa, = matris and matrem, = Sanskrit rndtaras and mdtaram. 
 Compare mater an and mayr with matrem and mere. The Armenians 
 have a word ovstr, " son," which possesses the Aryan termination 
 of -psi-ter, ma-£er, and daughter, and seems peculiar to the Arme- 
 nian language, where ovs means " teach," and also " shoulder." 
 
 t That ethe here means "when" or "if," was inferred by Dr. 
 Donaldson from its position, and without the guidance of any 
 linguistic resemblances. — Varronianus, p. 167 (2nd edition). This 
 gives force to those resemblances which I have adduced. 
 

 THE OLD ITALIANS. 103 
 
 Behistun yata^ yathd \ u -i „ 
 
 Sanskrit yadd J 
 
 12. Erai, "joyous," or " of joy." 
 Armenian erah, khrakh, ovrakh, " joyous, merry ": 
 
 ovrakh Uriel, " evfypaiveaOab" (Luke xv. 24) : ker, drb, 
 ev ovrakh ler, "(jzdye, irte, (fcal) evcfipalvov" (Luke xii. 19). 
 Armenian erakhan, " a banquet." Cf. epavos. 
 Erakhan would probably = erah khan, "joyous table," 
 as khan means " table" in Armenian, 
 
 13. Sie, "it be." 
 
 Armenian ize, "it be, it may be, may it be." A 
 comparison with the terminations of a Sanskrit 
 parasmaipada verb (infra, cap. v) tends to shew that 
 the Armenian has preserved here the precative form 
 of the substantive verb, and that the subjunctive 
 would be zie. 
 
 Sanskrit sydt : Latin sit (= siet) : German sei. 
 Italian sia. 
 
 14. Epana, " the feast/' nominative to sie. 
 Armenian eph, ephovmn, "cooking." Hebrew aphah, 
 
 "to cook." Latin epulari. Greek oTrrdte, eyjrco, o^frov. 
 For the termination of ep-ana, compare the Armenian 
 kap, "a bond," kap-el, "to contract," kap-an, "a 
 strait" — gel-el, "to roll," gel-an, "a cylinder" — 
 chah-el, "to expiate," chah-anay, "a priest." The - 
 Armenian prefers to terminate words with -ay, instead 
 of -a simply. Epana, and also kana (ante, p. 83), are 
 just like in form to the Sanskrit ddna, " gift." 
 
 15. Mi, "me." See 1. 
 
 16. Nethu, "of liquor." 
 
104 THE ASIATIC AFFINITIES OF 
 
 Armenian nivth, " substance, matter/' hetanivth, 
 " fluid substance, liquid/' which are declined like math 
 (6) , and thus give in the genitive, dative, and ablative, 
 nivthoy and hetanivthoy. Nethuns is the Etruscan 
 form of Neptunus. As we have in Armenian ivt and 
 et, " oil/' and givt and get, " village/' nivth would be 
 very nearly neth, though the Armenian iv is usually pro- 
 nounced like the German u, and occasionally like the 
 English u in tune. Compare the river-names Neda 
 and Nith, and the Sanskrit nadi, " river." 
 
 17. Nadav, " the guest." 
 
 Armenian ...nsdeh ) cc , ~ „ 
 
 -d . „ rpj^ > " stranger, foreigner. 
 
 Jrersian nasta* J ° ' ° 
 
 Arabic nazil, " stranger, foreigner, visitor, 
 
 Hebrew nasa , "to migrate"; wasa, "to err." 
 
 # C. Calidius Nasta, " Strange, Guest/' appears as a proper 
 name in a Neapolitan inscription (Donate p. 4), and Nastes is men- 
 tioned by Homer as one of the two Carian leaders. I have already 
 noticed (ante, p. 85) that Mestles was an Etruscan proper name. 
 It is a singular coincidence that the names, Mestles and Nasta, 
 should be found in Italian inscriptions, one of them at Volterra, 
 and that Homer, " Mseonii carminis ales/' who would probably 
 know what proper names were used in Lydia and Caria, should 
 have written : — 
 
 Mrfoaiv av ME20AH2 re kcli "Avrupos fjyrjadadrip. 
 and — 
 
 NA2TH2 ad Kapcov rjyfjaaro fiap^apofpca^cop. 
 Lethe and Lethi, again, were Etruscan proper names (ante, p. 59); 
 and Homer says that the Pelasgians at Troy were commanded by 
 Hippothous and Pylseus, 
 
 vie dvu) AH0OIO UeAaayov Tevra/midao. 
 In Tuschi, leth- means "kainpfen, drohen, schelten." 
 
 ' 
 
THE OLD ITALIANS. 105 
 
 Sanskrit nes, "ire, se movere." Nts-deh would 
 
 be one who goes from his 
 country (deh). 
 
 18. Helephu, "empties." 
 
 Czeiovl, "to pour"; zetkh, "drunken."* 
 Armenian . ■< lieiovl, "to pour out, i^eyelv^ (Rev.xvi.2). 
 \ hetov, " he pours out, he empties." 
 
 The root is liet, " pouring, flowing," which is found 
 just above in hetanivth (16), where nethu, being qualified 
 by helephu, acquires the meaning of hetanivth instead 
 of nivth. The formation of helephu from a root hel 
 may be thus illustrated from the Armenian : — 
 
 (sos, " causing tremor" (root). 
 < sos-aph-il, " to tremble." 
 ( thoth-oph-el, "to shake" (active). 
 
 ded-ev~el, "to reel." 
 
 J Jehovs-el, " to fly." 
 
 ( khovs-aph-el, " to fly." 
 
 7 [- " a trembling." 
 sarsaph J & 
 
 \ sarzil ^ 
 
 sarsil ^ " to tremble." 
 
 jsarsaphil J 
 
 Similar forms are presented by sosaphel, "to touch," 
 hachavel, "to dance," and sovthaphel, "to hasten": so 
 that hetaphov, as well as hetov, " he empties," might 
 exist in an Armenian dialect. 
 
 Helephu is the last word in the Inscription of Cervetri. 
 If all the Armenian words cited to explain this in- 
 
 * Cf. Thracian fcTAa, " ohos." — Botticher's Arica, p. 50. 
 
106 
 
 THE ASIATIC AFFINITIES OF 
 
 scription be now collected together, arid written in 
 Etruscan letters, we should obtain in grammatical 
 syntax, though the idiom might not be perfectly correct, 
 the following Armenian couplet : — 
 
 es mi hetu na es mathn mar em lezui tapean : 
 ethe erah ize ephumn, zis nithu nesteh helu. 
 
 Or, adopting such modifications as are warranted by 
 the Armenian language itself- — 
 
 me mi hetu na me mathu mar em lezui tapurini : 
 etlie erah zie ephanay, me nithu nesteh helaphu. 
 
 This distich differs but slightly from the Inscription 
 of Cervetri, and almost entirely preserves the metre in 
 which it is composed. Nor can it be said that the 
 sense which the Armenian supplies for the Etruscan is 
 at all forced or inappropriate ; but, on the contrary, 
 that it expresses exceedingly well what so festive a 
 nation might have inscribed on one of their drinking- 
 cups. For the meaning of the two verses would be:- — 
 
 1. 
 
 
 2. 
 
 
 mi 
 
 I 
 
 ethe 
 
 if 
 
 ni 
 
 not 
 
 erai 
 
 joyous 
 
 Ti-ethu 
 
 of water 
 
 sie 
 
 be 
 
 ma 
 
 but 
 
 epana, 
 
 the feast, 
 
 mi 
 
 I 
 
 mi 
 
 me 
 
 mathu 
 
 of wine 
 
 nethu 
 
 of liquor 
 
 mar 
 
 a cup 
 
 nastav 
 
 the guest 
 
 am 
 
 am 
 
 helephu. 
 
 empties. 
 
 Usiai 
 
 for the tongue 
 
 
 
 thipurenai: 
 
 thirsty : 
 
 
 
 
 The sentiment of the second verse brings to mind 
 
THE OLD ITALIANS. 107 
 
 Goethe's line on the drinking-cup of the King of Thule: 
 Er leerV ihnjeden Schmans. Some unimportant modi- 
 fications might be suggested in the interpretation of 
 the inscription. Thus lisiai thipurewai might be made 
 a locative, " on (his) thirsty tongue/' and connected 
 with the second verse instead of the first. If, again, 
 remembering that the Etruscan is several centuries 
 older than the earliest existing Armenian, we compare 
 the Etruscan Jcethu, mathu, and nethu, with the Sans- 
 krit madhu, the Zend mathu, the Greek fi&Ov, the 
 Phrygian /3e8v, "water," and the Macedonian fteOv, 
 " air" (both these last words being = Armenian vivth, - 
 "water, moisture, element, matter"), such analogies 
 would lead us to consider the Etruscan words as nomi- 
 natives or accusatives, rather than as genitives or 
 ablatives. If they be in the accusative, then we should 
 probably read, with Lepsius and Donaldson, mar am 
 instead of maram; and have to regard mar am as a 
 transitive verb of the second Armenian conjugation, 
 like tarn, "I give," or ertham, "I go," and signifying 
 "I contain," or "I dispense." Compare the Albanian 
 marr, " I contain," and the German fass and fassen. 
 This alteration would have the advantage of obviating 
 one little objection: for, if the Etruscan sie signify "it 
 be," "I am" should rather be em than am. We have, 
 however, both am and is, wast and ivert, in English, 
 where there is a similar change of vowel. If nethu be 
 a nominative or accusative, mi nethu would be ren- 
 dered "my liquor," or "my contents," mi being 
 equivalent to "my" or "of me," both rendered in 
 
108 THE ASIATIC AFFINITIES OF 
 
 Armenian by im. Finally, if mi nethu be a nominative, 
 hel&phu would have a passive or neuter signification — 
 the Armenian zetov is both active and neuter, like 
 " pours" in English — and nastav would be in the in- 
 strumental case, and = Armenian nsdehiv. The in- 
 strumental cases of nav, "a ship," and hin, "a woman," 
 namely navav and hiav, come still nearer in form to 
 nastav. 
 
 With these modifications the interpretation of the 
 inscription would be : " I do not contain water, but 
 wine: when there is a joyous feast, my liquor is poured 
 out by the guest on (his) thirsty tongue." Perhaps 
 this is on the whole the preferable interpretation of 
 the two. 
 
 It does not require the knowledge of many sen- 
 tences, nor of a large number of words and inflexions, 
 to enable us to pronounce upon the character of 
 any language ; so that the properties of the Etruscan 
 have probably been sufficiently displayed in the speci- 
 mens already given and analysed, which seem to 
 include all the forms whose meaning is tolerably 
 clear. The result is that, instead of there being no 
 language which can claim kindred with the Etrus- 
 can, there are, on the contrary, two in Asia which 
 may succeed in establishing a near relationship to 
 it by explaining it to a considerable extent. The 
 Armenian appears to do this in a very close manner, 
 especially when it is considered that Armenia and 
 Etruria are at opposite extremities of a long and 
 
THE OLD ITALIANS. 109 
 
 not entirely unbroken chain of old Thracian coun- 
 tries, like Wallachia and Portugal among those of 
 Latin speech. It may even be said, perhaps, that 
 the Armenian resembles the distant Etruscan more 
 than it does the neighbouring Phrygian, with which 
 it was connected by the ancients. But even if the 
 Armenian had perished with the rest of the Thra- 
 cian languages, of which only a few relics survive, 
 vet the affinities between the Sanskrit and the Etrus- 
 can would still have afforded some clue to indicate 
 who the Etruscans were. Their language would be 
 nearly allied to the Sanskrit, but would neverthe- 
 less belong to a different Aryan family, as the letter- 
 changes would imply. No doubt the Sanskrit has 
 some few advantages over the Armenian in the com- 
 parison of languages. Thus the Etruscan semph-, 
 r seven," is nearer to the Sanskrit saptan than it is to 
 the Zend liaptan, the Persian haft, or the Armenian 
 evthn; and the Etruscan sas, " six," is nearer to the 
 Sanskrit sas than it is to the Armenian wez. But 
 the Latin septem and sex are likewise nearer to the 
 Sanskrit saptan and sas than they are to the Greek 
 eirrd and e£ ; and yet the Latin and Greek are con- 
 sidered to belong to the same Aryan family of lan- 
 guages, while the Sanskrit and Latin are not so classed 
 together. There are letter-changes which distinguish 
 one Aryan family from another, as there are letter- 
 changes which distinguish different members of the 
 same family from one another. There is, besides, no 
 letter-change in the case of the Armenian wez and the 
 
110 THE ASIATIC AFFINITIES OF 
 
 Etruscan sas, as botli would be ultimately derived, 
 along with the Albanian gyas(te) and the Persian sas, 
 from a form like the Zend Ichsvas, by the omission of 
 different letters, after the manner in which the Old 
 Norse fimm and the Anglo-Saxon fif are deduced 
 from the Gothic form, fimf. The argument from simi- 
 larity or dissimilarity of numerals must not be pressed 
 too far. Thus the Swedish tio and the German zehn 
 have not one letter in common. The Gothic fidvor, 
 too, resembles the Welsh pedwar more than the Ger- 
 man vier : yet the Gothic was a Teutonic, not a Celtic 
 dialect. The right conclusion would be, that the 
 Gothic and Welsh forms are older than the German, 
 as the Zend thri and the Etruscan thr- are older than 
 the Armenian ere, " three." So, again, the Welsh 
 pump and the Breton pemp are more like to the Gothic 
 fimf than they are to the Gaelic cuig, as the Welsh 
 pedwar and the Breton pevar are more like to the 
 Gothic fidvor and the Anglo-Saxon feover than they 
 are to the Gaelic ceathair* Nor is the advantage all 
 on the side of the Sanskrit in respect of the Etruscan 
 numerals. Mack (with me- and muv-), " one," is 
 Armenian, but not Sanskrit ; and the Armenian king, 
 "five," leads us from the Sanskrit pancan to the 
 Etruscan Idem or chiem. 
 
 It might be conjectured, on account of proximity, 
 
 * To get from pemp to cuig, we should pass through the Greek 
 ire five and it€i*t€, the Lithuanian penhi, the Armenian King, and the 
 Latin quinque. Greek and Oscan resemble Welsh, as Latin re- 
 sembles Gaelic. 
 
THE OLD ITALIANS. Ill 
 
 that the Etruscans were Illyrians rather than Thracians, 
 if the Illyrians be supposed to constitute a distinct 
 Aryan family. But, even if we set aside other argu- 
 ments, and lay more stress than is allowable on numerals, 
 there would yet be no necessity for such a conclusion, 
 as the Albanian numerals hardly come nearer the 
 Etruscan than the Armenian numerals do. These are 
 the Albanian numerals from I to x : nye, du, ire, hater, 
 pese,* gyasftej, stafiej, teftej, nen(de) , dhyefte'J. The 
 corresponding Armenian numerals are : ez (and also 
 mi, mov, and men), erlwv (not Aryan), er or ere(ch), 
 cor(ch) or char, king, wez (in composition wes and 
 wath), evthn, ovth, inn (— inert), tasn. 
 
 Such advantages as the Sanskrit may have over the 
 Armenian in some few instances cannot counterbalance 
 the weight of evidence on the other side, so as to take 
 the Etruscans out of the Thracian family. It is not to 
 be expected that every Thracian language should be 
 quite like the Armenian, any more than that e^ery 
 Teutonic language should be quite like the English, or 
 that every Celtic language should be quite like the 
 Welsh, or every Neo-Latin language quite like the 
 French. And, while the Sanskrit explains so much of 
 the Etruscan, it almost, by that very fact, disposes of 
 its own claims to include the Etruscans in the Indian 
 family. Such a word as sutliina, for instance, if ex- 
 plained by the Sanskrit hit, " Diis offerre" — and a 
 word found singly on votive offerings is perfectly so 
 
 * Compare the Lettish, peezi, which belongs to the same family 
 as the Lithuanian penki, 
 
112 THE ASIATIC AFFINITIES OF 
 
 explained — is nearly decisive by itself. Suihina would 
 not be a Sanskrit word ; while, on the other hand, the 
 Armenian brings out suthina from hu by presenting 
 both the right letter-change, as well as the termina- 
 tion, in the word zohovthivn, " sacrifice." It enables 
 us also to form such Etruscan words as zilachnhe and 
 thij)iirenai from Aryan roots, where other Aryan lan- 
 guages would not qualify us to construct them \ and it 
 has, in addition, the Etruscan I terminations, besides 
 the singular Etruscan peculiarity of retaining in poetry 
 the vowel which is discarded in prose. The Slavonian 
 family of languages might compete with the Armenian 
 on the ground of the letter-changes, but would fall 
 far behind it, as well as behind the Sanskrit, in ex- 
 plaining Etruscan words. There is likewise a geo- 
 graphical improbability against the Sanskrit by reason 
 of distance, and because Armenia fills up the gap be- 
 tween the Caucasian and Semitic nations. 
 
 The evidence in favour of the Armenian affinities of 
 the Etruscan is not exhausted by the Etruscan in- 
 scriptions. For we find in Etruria place-names re- 
 sembling the Armenian sen and lori, which may be 
 described as the town and home, or -ton and -ham, of 
 Armenia; as we find among Dacian plant-names terms 
 like the Armenian Jchot and det, which are the kraut 
 and wurz of Armenia. Finally, the Etruscan and 
 Rhaetian were said on sufficient authority to be cog- 
 nate languages, and in Rhaetia there are still apparent 
 relics of an Armenian dialect ; while in the Pyrenean 
 sern-eille, " glacier," a similar dialect seems to have 
 
THE OLD ITALIANS. 113 
 
 penetrated still farther west than Etruria. It can 
 hardly be accidental that the only Thracian language 
 still existing should emerge wherever the ancients 
 have placed a Thracian people.* 
 
 * The quarter in which to look for the right language to explain 
 the Etruscan was indicated by Bonarruoti a century and a half ago: 
 (t Hortari postremo fas mihi sit, doctos prsecipue Unguis Orientalibus 
 viros, ut animi vires intendant, ad illustrandam veterem Etruscam 
 linguam, tot jam seculis deperditam. Et quis vetat sperare, quod 
 temporum decursu emergat aliquis, qui dimcilem et inaccessam 
 viani aperiat, et penetralia linguss hujus reseret ?" Mebuhr was 
 less sanguine in his expectations. " People," he says, " feel an ex- 
 traordinary curiosity to discover the Etruscan language ; and who 
 would not entertain this sentiment ? I would give a considerable 
 part of my worldly means as a prize, if it were discovered ; for an 
 entirely new light would then be spread over the ethnography of 
 ancient Italy. But, however desirable it may be, it does not follow 
 that the thing is attainable." And yet it has been known from 
 the first that the ages of deceased persons were denoted in Etruscan 
 by such forms as avil ril lxv, ril leine lv, and lupu evils xvn, which 
 might, it would seem, have opened the way to the discovery, as 
 they supply us with four words whose meaning can hardly be said 
 to be doubtful, and which are thoroughly explained in every respect 
 by the Armenian and Sanskrit languages. 
 
114 THE ASIATIC AFFINITIES OF 
 
 CHAPTER V. 
 
 CONJUGATIONS AND NUMERALS, 
 
 It may be considered unnecessary to prove the Aryan 
 character of the Armenian language by an analysis of 
 the Armenian verb ; so that the evidence in demon- 
 stration of the Etruscan being an Aryan language of 
 the Armenian or Thracian type might have been closed 
 with the interpretation of the Inscription of Cervetri. 
 But the Armenians are even yet not universally ad- 
 mitted into the Aryan family, although it is difficult to 
 perceive on what grounds their right to such ad- 
 mission has been disputed ; as their vocabulary, and, 
 what is of more importance, their grammar also, are 
 both decidedly Aryan. That their vocabulary is so in 
 substance, the previous chapters may have sufficiently 
 shewn ; and in the present chapter I shall endeavour, 
 by an examination of Armenian conjugations, to com- 
 plete the proof that their grammar is so too. Albanian 
 and Rhgeto-Romance conjugations will likewise be 
 found compared with similar forms in other Aryan 
 languages : and from these I have passed to Caucasian 
 and Basque conjugations, in order to exemplify a 
 little how far these languages deviate from the Aryan, 
 and approach or differ from one another. Lastly, as 
 
THE OLD ITALIANS. 
 
 115 
 
 the primeval population of Europe and Asia Minor 
 may be conjectured (ante, p. 12) to have been com- 
 posed of Caucasian, Basque, and Finnish elements, I 
 have attempted to gain some insight into the obscure 
 question of the relationship among these three races 
 by an examination of their numerals. There are some 
 indications of primeval affinity here which I have not 
 found noticed, and which may appropriately bring 
 this present inquiry to a termination. 
 
 Aryan Conjugations. 
 Present Indicative. 
 
 Lithuanian. 
 
 Sanskrit. Rh.83 to- Romance. 
 
 esmi 
 
 
 asmi 
 
 stmt 
 
 essi 
 
 
 asi 
 
 eis 
 
 esti 
 
 
 asti 
 
 ei 
 
 esma 
 
 
 smas 
 
 essen 
 
 este 
 
 
 stha 
 
 esses 
 
 esti 
 
 
 santi 
 
 ean 
 
 Albanian. 
 
 Zend. 
 
 Behistun. 
 
 Armenian. 
 
 y am 
 
 ahmi 
 
 amiya 
 
 em 
 
 ye 
 
 alii 
 
 
 es 
 
 este 
 
 asti 
 
 astiya 
 
 $ 
 
 yemi 
 
 hmahi 
 
 amahya 
 
 emch* 
 
 yini 
 
 stha 
 
 
 ech 
 
 ydne 
 
 henti 
 
 
 en 
 
 # Here the Sanskrit -mas is converted into the Zend and Behis- 
 tun -mail-, and the Armenian -mech, as we found (ante, p. 44 ) the 
 Sanskrit mas, " moon," converted into the Behistun m&h-, the Ar- 
 menian mall-, and the Etruscan mach. 
 
116 
 
 THE ASIATIC AFFINITIES OF 
 
 French. 
 
 Latin. 
 
 
 Italian. 
 
 suis 
 
 sum 
 
 
 sono 
 
 es 
 
 es 
 
 
 sei 
 
 est 
 
 est 
 
 
 e 
 
 sommes siimns 
 
 
 siamo 
 
 etes 
 
 est is 
 
 
 siete 
 
 sont 
 
 sunt 
 
 • 
 
 sono* 
 
 
 Imperfect Indicative. 
 
 
 Sanskrit 
 
 S anskri t . Alb anian 
 
 . Armenian 
 
 (1st aorist). 
 
 
 
 (1st aorist). 
 
 adik-sam 
 
 as am 
 
 yese 
 
 sir-ezi 
 
 -sas 
 
 asis 
 
 yese 
 
 -ezer 
 
 -sat 
 
 CLSlt 
 
 is 
 
 -eaz 
 
 -soma 
 
 asm a 
 
 yesem 
 
 -ezachf 
 
 -sata 
 
 dsta 
 
 yesete 
 
 -ezich 
 
 -san 
 
 dsan 
 
 isne 
 
 -ezin 
 
 Zend. 
 
 Behistun. 
 
 Armenian. 
 
 
 dham 
 
 
 ei 
 
 eir 
 
 as 
 
 alia 
 
 
 er 
 
 each 
 
 eich 
 
 
 alia 
 
 
 ein 
 
 * Dante on two occasions uses en and enno for sono ; forms 
 which are like the Armenian en and the Rhseto-Romance ean. 
 
 f Here the Armenian omits the m of the Sanskrit and Albanian, 
 but retains, under the form of ch, the final s which they drop. So, 
 in the 2nd pers. piur., t is dropped, but s retained, in Armenian. 
 
THE OLD ITALIANS. 
 
 117 
 
 Sanskrit. 
 
 Zend. 
 
 Behistun 
 
 abliavam 
 
 baom 
 
 abavam 
 
 abhavas 
 
 bavd 
 
 
 abhavat 
 
 bavat 
 
 abava 
 
 abhavdma 
 
 bavdma 
 
 
 abhavata 
 
 bavata 
 
 
 abhavan 
 
 
 abava 
 
 3to-Romance. 
 
 
 Albanian (aorist). 
 
 fova 
 
 
 Jcerko-va* 
 
 fovas 
 
 
 -ve 
 
 fova 
 
 
 -i 
 
 fovan 
 
 
 -nam 
 
 fovas 
 
 
 -uate 
 
 fovan 
 
 
 -uane 
 
 The root bhu or fu is not found in the Armenian 
 language, which employs U or et instead. This may 
 account for there being no such words as fnius or 
 puia in Armenian (ante, p. 60). In the Armenian 
 first aorist, as in sir-ezi, "1 loved/' the s of the Sans- 
 krit root as appears as z, as it does also in the Arme- 
 nian subjunctive. Scrip-si, (i)^>i\-r>cra (= icjyiXe-eaa), 
 and sir-ezi, are analogous forms, all bearing a similar 
 relation to the imperfect or preterite of the substan- 
 tive verb that the corresponding form in Sanskrit 
 does : and -ezi is to ei what -eaa is to rjv, as may be 
 readily seen when the Greek and Armenian forms are 
 thus placed together : — 
 
 * Kerkova, " I sought," s^ Italian cercava, " I was seeking." 
 
113 
 
 THE ASIATIC AFFINITIES OF 
 
 
 Greek. 
 
 
 Armenian. 
 
 
 1st aorist terminations. 
 
 Impft. 
 
 1st aorist terminations. 
 
 Impft. 
 
 -eaa 
 
 -cnx 
 
 rjv 
 
 -ezi 
 
 Si 
 
 -eaa$ 
 
 -acts 
 
 v$ 
 
 -ezer 
 
 iir 
 
 -ecre 
 
 -(76 
 
 V 
 
 -eaz 
 
 A 
 
 er 
 
 -eaafiev 
 
 -aa[Mev 
 
 rjfxev 
 
 -ezach 
 
 Sack 
 
 -cadre 
 
 -a are 
 
 rjre 
 
 -ezich 
 
 dich 
 
 -ecrav 
 
 -aav 
 
 rjorav 
 
 -ezin 
 
 Sin 
 
 In the imperfect of the substantive verb, both lan- 
 guages assume the augment (which lengthens the 
 initial e), but drop, with one exception in each case, 
 the sibilant of the root, which the Greek retains in the 
 3rd pers. plur., and the Armenian, under the form of 
 r> in the 3rd pers. sing. The 2nd aorist, as well as 
 the 1st, is formed from the imperfect or preterite of 
 the substantive verb ; and the manner in which it is 
 done is again similar in Sanskrit, Greek, and Arme- 
 nian ; as may be exemplified by the 2nd aorist of " to 
 place" in those three languages, to which a second 
 form of the Albanian aorist is added : — 
 
 
 2nd Aorist. 
 
 
 Sanskrit. 
 
 Greek. 
 
 Armenian. 
 
 Albanian. 
 
 a-dh-dm 
 
 6-6-r\v 
 
 e-d-i 
 
 plyah-a 
 
 a-dh-ds 
 
 e-0-rj? 
 
 e-d-er 
 
 plyak-e 
 
 a-dh-dt 
 
 6-6-7) 
 
 e-d 
 
 plyah-i 
 
 a-dh-dm a 
 
 6-6-6jbL6V 
 
 e-d-ach 
 
 plyak-m 
 
 a-dh-dta 
 
 6-9 -6T€ 
 
 e-d-ich 
 
 plyah-te 
 
 a-djh-us 
 
 6-6 -6(7 av 
 
 e-d-in* 
 
 plyak-ne 
 
 * Though there are two forms of the aorist in Armenian, yet no 
 verb has more than one of them, except in the participle. 
 
THE OLD ITALIANS. 
 
 119 
 
 Sir-Si, cc amabam," sir-ezi, " amavi," and e-d-i, 
 " posui," exhibit the three preterite forms of the Ar- 
 menian. 
 
 Present Subjunctive, Potential, & Precative or Optative. 
 Sansk. (subj.)** Khasto-Kom. (subj., pot. prec.) 
 
 syam 
 
 
 seig 
 
 syas 
 
 
 seias 
 
 syat 
 
 
 seig 
 
 sydma 
 
 
 seian 
 
 syata 
 
 
 seias 
 
 syan 
 
 
 seian 
 
 Sansk. (pot.) 
 
 Alb, 
 
 (subj., pot., prec.) 
 
 yam 
 
 
 yem 
 
 yds 
 
 
 A' 
 
 yes 
 
 ydt 
 
 
 yit 
 
 ydma 
 
 
 yemi 
 
 ydta 
 
 
 yini 
 
 yus 
 
 
 yene 
 
 Sansk. (prec.) 
 
 Arm. 
 
 (subj., pot., prec.) 
 
 ydsam 
 
 
 izem 
 
 yds 
 
 
 izes 
 
 ydt 
 
 
 izS 
 
 ydsma 
 
 
 izemch 
 
 ydsta 
 
 
 izech 
 
 ydsus 
 
 
 izen 
 
 If we omit the z 
 
 in these Armenian forms, we get 
 
 the Sanskrit and Albanian potential ; and if we trans- 
 
 * This and the two following forms are taken from a parasmai- 
 pada verb. 
 
120 
 
 THE ASTxlTIC AFFINITIES OF 
 
 pose the i and z, the Sanskrit and Latin subjunctive, 
 for sim = stem. The Armenian equivalent to the Latin 
 sit would thus be zie, = Etruscan sie [ante, p. 103). 
 
 The only Armenian future is a futnrum exactum, 
 like scripsero and rvfyOrjaoiuu, being formed from the 
 aorist by the addition of terminations which are modi- 
 fications of the different persons of the subjunctive of 
 the substantive verb; as sirez-iz, "amabo" (amavero), 
 from sirez-i, "amavi." One of the following examples 
 will exhibit a first aorist future of the e conjugation, 
 and the other a second aorist future of the i conjuga- 
 tion: an Albanian aorist subjunctive is added, as being 
 a form almost identical with the fiuturum exactum : — 
 
 Subjunctive. 
 
 sir-izem* 
 
 sir -izes 
 
 sir-izS 
 
 sir-izemch 
 
 sir-izech 
 
 siv-izm 
 
 sirez-iz 
 
 sires-zes 
 &ires-ze 
 sires-zovch 
 sires-ffich 
 
 sires-zen 
 
 Future. 
 
 lin-izim 
 
 lin-izis 
 
 lin-izi 
 
 \m-izimch 
 
 lin-izich 
 
 lin-izin 
 
 If-izim 
 
 l-izis 
 
 l-izi 
 
 \-izOVcll% 
 
 l-izich, or -igich 
 l-izin 
 
 * The subjunctive of the substantive verb is accurately pre- 
 served here throughout all the persons. 
 
 + The Armenian, like the Sanskrit and Greek, drops the con- 
 jugational n in the aorist and future. 
 
 £ Cf. Armenian anovn = nomen. 
 
THE OLD ITALIANS. 121 
 
 Albanian Aorist Subjunctive. 
 
 plyak-sa or plyak-^m 
 
 plyak-s or plyak-£s 
 
 plyak-fe" 
 
 plyak-sim or plyak-tsim, 
 
 plyak-si or iplyeik-tsi 
 
 plyak-smef or plyak-^sme. 
 The Albanian ts, in the second of these forms, shews a 
 tendency to convert the Sanskrit s into a sound re- 
 sembling the Armenian z (tz) ovg(ds). In the Ar- 
 menian sires zes, etc., the z of sirezi becomes s. The 
 Armenian and Albanian forms are both analogous in 
 their terminations to the Sanskrit 2nd future : — 
 
 Sanskrit. 
 
 Subj. of "to be." Subj. terminations. 2nd Fut. terminations. 
 
 sydm syam -sydmi 
 
 syds -syas -syasi 
 
 syat -syat -syati 
 
 sydma -sydma -sydmas 
 
 syat a -syata -syatha 
 
 syns -syan -syanti 
 
 The Armenian perfect is formed by combining the 
 
 preterite participle with the present indicative of the 
 
 substantive verb. For the participle compare :— 
 
 A . f sir-eal, " loved/' or " having loved." 
 
 Armenian ^ pah _ eal ^ « kept ;> or « having kept .» 
 
 Old Slavonic by-t', " having been." 
 
 Mahratti pdh-ild, " seen" (cf. Daci.an $i6o<j>6e6-eka; 
 
 ante, p. 9). 
 
 And, for the perfect, compare : — 
 
122 THE ASIATIC AFFINITIES OF 
 
 Armenian sireal em, " I have loved" (root sir) . 
 
 Bengali lidrihim, " I made" (root kor)* 
 
 Bohemian byl sent } u lw ^> (root b) 
 Polish by tern J l WaS ^ r00t by) ' 
 
 An Armenian pluperfect, like sireal ei, "I had loved/' 
 i. e. " I was having loved/' calls for no observation ; 
 and the same may be said of the imperative : — 
 
 Sanskrit. Latin. Armenian. 
 
 edhi es > er 
 
 sta este tick or erovck* 
 
 Caucasian and Basque Conjugations. 
 
 Basque verbs are usually conjugated by combining 
 
 a few auxiliary verbs and pronouns, united together in 
 
 various agglutinate or incorporated forms, with the 
 
 present participle, the preterite participle, and the 
 
 future participle, of a particular verb. These participles 
 
 are sometimes called infinitives. The Armenian sireal 
 
 em and the French aimer-ai are modes of conjugation 
 
 like those in Basque. Though Basque verbs have a 
 
 strange appearance on account of the extent to which 
 
 agglutination or incorporation is carried, yet they are 
 
 simple enough when analysed. Thus ecarri nezaque, 
 
 "je pouvais apporter," is quite plain when resolved 
 
 into ecarri n-eza-que, " to-carry I-was-able"; n- being 
 
 = ni, " I," eza the substantive verb, and que = Latin 
 
 que-o. Nezaque is only " poteram" with the order of 
 
 * These analogies are derived from Bopp (V. G., p. 1159). His 
 argument, that korildm "von participialem Ursprung zu sein 
 scheint," would be strengthened by the Armenian sireal em, which 
 is not agglutinate like horil<lm. 
 
THE OLD ITALIANS. 128 
 
 the three elements reversed, as in the English " I was 
 able." So again, ecarri guinitzaiztzuque, "nous te les 
 pouvions apporter," is ecarri guin-itz-aiz-tzu-que, " to- 
 earry we-them-were-to you-able." 
 
 As personal pronouns present some of the most 
 intimate signs of affinity between different languages, 
 I shall endeavour to detach from these agglutinate 
 forms the personal pronouns, or pronominal affixes, 
 which are the subjects of the auxiliary verbs. It will 
 be found that there is more than one such affix for 
 each person, and that the Basque pronouns, in their 
 present state, cannot explain several of the affixes. In 
 the forms which I shall cite, all but the subjective 
 affixes will be enclosed in brackets, so as to leave 
 nothing but what belongs to the pronouns implicitly 
 found in the Basque language ; and I shall begin by 
 placing the Basque naiz, "I am/' between the Georgian 
 var } "I am," and machus, "I have." In both languages, 
 as will be seen, the root of "being" has a common 
 origin with the Aryan root ; and this root takes three 
 forms in Basque, as it does in the English am, art, is, 
 ar(e). 
 
 Georgian. Basque. Georgian. 
 
 v(ar) n(aiz) m (achus) 
 
 hh(ar) (aiz), c(era) g (achus) 
 
 (ar)s d(a) (achus, achun) 
 
 v(ar)th gu{era) gv(achus) 
 
 Jch(ar)th c(era)te g (achus) th 
 
 ( ar)ian d (ira) h (chon) ian 
 
 There is a very clear resemblance here between 
 
124 THE ASIATIC AFFINITIES OF 
 
 the Basque and the Georgian in the 2nd pers. sing, 
 and in the 1st and 2nd pers. plur. Four other Basque 
 conjugations are : — 
 
 n(uen) n(uque) 
 
 (nen), cencl(uen) cend(uque) 
 
 z(uen), ce(uen) l(uque) 
 
 gnen(uen), gaend(uen)* guend(aque) 
 
 cen(u)t(en), cend(u)t(en) cend(uque)te 
 
 z(u)t(en), cenQiien) l{uque)te 
 
 n(ezan) n(ioteque) 
 
 cen(ezan) cen(ioteque) 
 
 c(ezan) l(ioteque) 
 
 guen(ezan) gain (ioteque) 
 
 cen(eza)te(n) ciniioteque) 
 
 c(eza)te(n) l(ioteque) 
 
 Z-t(e), l-te, cend-te, and cen-te mark the plural of the 
 persons z and Z, " he/' and cend and cen, " thou/' just 
 as we previously found in Basque, c(era), " thou art/" 
 and c(era)te, "ye are"; and in Georgian, v{cur), "I am/' 
 and v(ar)th, "we are"; Wi(ar), "thou art/ 5 and kh(ar)th, 
 "ye are."f In the Basque verbs cited above, the pro- 
 nominal affixes were prefixed : in the verb which follows, 
 they are post&xed : — 
 
 * The d in guend- and cend- seems not radical, but phonetic or 
 euphonic. The sound of d rises between n and u, like that of p 
 between m and s, as in Sampson. 
 
 f Compare Ossetic forms like lag, " man," lagthii, " men"; ye, 
 "he," yetha, "they." The Ossetic plural suffix is -th'ti or -the 
 (Sjogren, p. 52). 
 
THE OLD ITALIANS. 
 
 125 
 
 (de)t 
 
 (de)c, (de)n } de(zu) 
 
 (de)u 
 
 (de)g% 
 
 (de)zue, (de)zute 
 
 (de)ue, (d)ute 
 These pronominal affixes,, with the addition of the 
 actual Basque pronouns, will give us the pronominal 
 forms contained in Basque for " X, thou, he, we, ye, 
 they." But, before they are compared with Caucasian 
 and other forms, it is advisable to mention the com- 
 plete Georgian form for "I." The pronoun itself is me, 
 but the genitive cemi } as well as the other cases of the 
 pronoun, shew that cem is another or a more perfect 
 form, just as sen, "thou," makes seni, "of thee." The 
 Aryan analogies to the Basque will suggest themselves 
 without notice. The Basque pronominal forms are : — 
 
 ni \ Hungarian en; Suanian noi 3 "we, nos"; 
 
 n- ) Lesgi nise, "we." 
 
 -1 Tuschi so, -s. 
 
 Ten or teni may be the complete Basque form. 
 
 Lesgi den, tun;* Georgian cem(i) ; Lazic 
 skim(i). 
 
 r hi. Tuschi ho, ~h. 
 
 Suanian si. Greek av. 
 Georgian ~kh-, g-. 
 \ Georgian sen ; Turkish sen ; 
 j Lazic sJcan(i)> 
 The complete Basque form might be 
 zchuen. 
 
 "L"i 
 
 I 
 
 " Thou." ^ 
 
 zu, 
 
 -&u. 
 
 <S 
 
 •6. 
 
 -n 
 
 
 cen- 
 
 •, cm 
 
 * There are several Lesgi dialects. My authority for them is 
 Klaproth/s Kaukasische Sprachen. 
 
120 
 
 THE ASIATIC AFFINITIES OF 
 
 a, au, o \ Tuschi o; Turkish o; 
 -u ) khasian ui. 
 
 Ab- 
 
 ce-, c-. Georgian igi ; Lesgi hey en, 
 
 gen. hegei. 
 
 j z-. Turkish su. 
 " He, this. that. < 7 m i • i 7 T • •? a 
 
 3 ' W-. lurkish oi ; Lesgi %l ; Suanian 
 
 alle. 
 d-. Bsthonianto; Lappouic ta(t). 
 There would be more than a single 
 
 form here. 
 
 r gu, gu- } -gu. Georgian gv-; Tuschi wai, thcho; 
 Lazic sku. 
 u We."-<( guen-, guin-. Georgian even ; Lazic shun(i). 
 Guen seems plainly the complete Basque 
 form. 
 
 r zue(c), -zue, -zute. Tuschi su, -s. 
 c — te. Georgian hh — th, g — th. 
 <(y » J cin-, cen-, cen — te. Georgian thchven, Lazic 
 tqua, tquan(i). 
 The complete Basque form seems nearly 
 zcuen. 
 
 r aie(c), oleic), ~ue. Lesgi hai, hoi, ua, 
 " this." 
 -ute ^ 
 
 z — t > See under " he." 
 " They." \ c—te J 
 
 j J J- Turkish -l(er) ; Lesgi il, "he." 
 
 d-. Lapponic tah ; Lesgi ti. 
 ^cen~. Georgian igini ; Lazic hini. 
 
 Assuming ten, zcliuen, guen, and zcuen, as the com- ' 
 plete or primitive Basque forms for "I," "thou," "we," 
 and "ye," some suggestive comparisons may be made 
 between the Aryan, Caucasian, and Basque languages : — 
 
THE OLD ITALIANS. 
 
 127 
 
 Y "I," = ah(am). 
 
 i 
 J 
 
 >"Thou/' = tv(am). 
 
 Greek iycov ^ 
 
 Sanskrit . ..aham \ 
 
 Zend azem j 
 
 Behistun. ..adam 
 
 Lesgi den 
 
 Basque ...ten 
 Georgian . cem 
 Lazic shim 
 
 Zend turn ^ 
 
 Behistun . . . t'huwam \ 
 Sanskrit ...tvam 
 Basque ...zchuen 
 
 Lazic shorn 
 
 Georgian... sen 
 
 Tuschi wai 
 
 Behistun . . .way am 
 
 Sanskrit . . .vayam 
 
 Zend vaem 
 
 Basque . .guen 
 
 Georgian. ..even 
 
 T . ( shun 
 
 Jjazic ... < / 7 
 ( sku 
 
 Tuschi...... thcho 
 
 Sanskrit . . .yuyam 
 Zend yuzem 
 
 Basque ...zeicen 
 
 Georgian. ..thchven 
 
 T . f tquan 
 Lazic . -! / 
 
 ( tqua 
 
 Welsh chivi • 
 
 Tuschi su J 
 
 Although the Caucasian and Basque languages are 
 
 far from being Aryan., yet it seems as if there were 
 
 some ancient connexion between the three forms of 
 
 speech. There may have been some group of men 
 
 in Western Asia, from which the Basques first broke 
 
 >- "We," = vay(am). 
 
 l 
 
 i 
 
 ! 
 i 
 y " Ye/' = yuy(am). 
 
128 THE ASIATIC AFFINITIES OF 
 
 off, and then the Caucasians ; while the remainder, or 
 at least a part of the remainder, subsequently moulded 
 their language into the primitive Aryan, which be- 
 came subject to different modifications when the Aryan 
 race spread abroad, and became divided into families, 
 and subdivided into nations. For it is with the most 
 ancient Aryan forms that the Georgian and Basque 
 languages appear connected by their pronouns. The 
 Georgian th-chve-n and the early Basque z-cue-n, "ye" 
 must be older than the Welsh chwi, " ye," if allied to 
 it; and the Georgian c-ven and the Basque g-uen seem 
 even more ancient than the Sanskrit vayam, "we," 
 though the Tuschi ivai would be a younger form. If 
 the resemblances in the cases of " we" and " ye" 
 justify us in identifying the Georgian and Basque ter- 
 mination -en with the Sanskrit termination -am, then 
 we should have a right to apply the same principle to 
 "I" and "thou." Here then the Caucasian and Basque 
 would retain signs of a characteristic which is only 
 found in the most ancient Aryan languages : " Den 
 ausgang -am in aham tvam, azem turn, entbehren alle 
 jiingeren sprachen" (Grimm) . Tet we find in Cauca- 
 sian dialects den, cem[i), and sldm(i), — sen and shan(i), 
 for the singular of the first two personal pronouns, 
 and can construct from the Basque, ten, " I," and 
 zchnen, " thou." 
 
 One more resemblance between the Caucasian and 
 the Basque is worth notice. It may be seen from such 
 forms as jango nu-gue, " que je mangerais," when 
 compared with ecarri neza-que, "je pouvais apporter/' 
 
THE OLD ITALIANS. 129 
 
 that the Basque conditional or potential is formed by 
 the suffix que, implying " ability/ 5 The subjunctive 
 is formed in a similar manner by the suffix Id. Thus 
 we have :— 
 
 venio/' nator natorreld, " veniam.^ 
 
 venis/' ator atorreld, " venias." 
 
 venit/' dator datorreld, " veniat." 
 
 veriimus/' gatoz gatorreld, u veniamus." 
 
 venitis/' zatozte zatoceld, " veniatis." 
 
 veniunt/' datoz datozteld, " veniant." 
 
 The subjunctive is formed exactly in the same waj r 
 in Tuschi by the suffix le, which is referred by Schief- 
 ner to the verb la(ar), " to wish/* Cf. \d~co. In 
 Tuschi, do is " facit." and dole is " faciat/' like as in 
 Basque dator is "venit/' and datorrela is "veniat." 
 The conditional in Tuschi is formed by the suffix lie 
 or h : as dahe from da, " he is/' and dolt from do, " he 
 does." In Lesgi we have bugo, " er ist/' and bugahi, 
 " es sei." 
 
 Numerals. 
 
 It has been said by Grimm, in the chapter of his 
 GeschicMe der Deittschen Sprache which is devoted to 
 the subject of original affinity (ur ver wands cliaft) : "Mit 
 recht hat man drei kennzeichen ermittelt, welche in 
 samtlichen urverwandten spracheix, wo nicht unveran- 
 dert, doch hochst deutlich und eigenthumlich an- 
 zutreffen sind, und fiiglich als symbol derselben auf- 
 gestellt werden diirfen. Ich meine die iibereinkunft 
 der zahlen, personlichen pronomina, und einzelner 
 
 K 
 
130 
 
 THE ASIATIC AFFINITIES OF 
 
 formen des substantiven verbums" Conjugations and 
 personal pronouns I have already examined as evi- 
 dences of early affinity ; and now, by the aid of such a 
 comparison and analysis of numerals as I am able to 
 make, I shall endeavour to penetrate a little further, 
 if possible, into the difficult subject of the original re- 
 lationship of the Basques, the Fins, and the Caucasians, 
 the three races by whom Europe was probably peopled 
 at the time when the Aryans first entered it. It will 
 not be necessary to set down Aryan numerals, as they 
 are so well known : the others which I shall notice are 
 these : — 
 
 
 I 
 
 II 
 
 in 
 
 IV 
 
 v. 
 
 
 
 Finnish. 
 
 
 
 Fin 
 
 ..yksi 
 
 kaksi 
 
 kohni 
 
 nelja 
 
 wiisi. 
 
 Esthonian . 
 
 . .uts* 
 
 katsf 
 
 holm 
 
 nelli 
 
 wiis. 
 
 Lapponic — 
 
 . . akt 
 
 qwehte 
 
 holm 
 
 nelje 
 
 wit. 
 
 ►Syrianic ... 
 
 ..otik 
 
 hyh 
 
 kujm 
 
 njolj 
 
 vit. 
 
 Hungarian . 
 
 ..egy 
 
 Icettd 
 
 Basqu 
 
 hdrom 
 e. 
 
 negy% 
 
 of. 
 
 
 bat 
 
 u 
 
 liiru 
 
 lau 
 
 bost. 
 
 
 
 Caucasian. 
 
 
 
 Georgian 
 
 ..erthi 
 
 ori§ 
 
 sami 
 
 othkhi 
 
 hliutlti. 
 
 Tuschi 
 
 ..zlia 
 
 si 
 
 qho 
 
 dhew 
 
 plichi. 
 
 Circassian . 
 
 ..se 
 
 tu 
 
 si 
 
 ptVe 
 
 t'chu. 
 
 Abkhasian . 
 
 ..aha 
 
 vi 
 
 khi 
 
 phsi 
 
 elm. 
 
 
 
 Turkish. 
 
 
 
 
 blr 
 
 iki 
 
 uc 
 
 diirt 
 
 bes. 
 
 * Genitive, iltte. f Genitive, katte. 
 
 X Compare these Finnish numerals for " Jour" with the Tamil 
 n'mgu, nO.lu, "four." 
 
 § Compare the Chinese <lr, "two." 
 
zhhra 
 
 athi. 
 
 iss 
 
 itt. 
 
 hgib 
 
 pse. 
 
 s 
 
 sva.f 
 
 THE OLD ITALIANS. 131 
 
 VI VII VIII IX X.* 
 
 Finnish. 
 
 Fin huusi seitsen hahdehsa yhclehsa hymraen. 
 
 Esthonian . huus seitse hattesa iittesa hiimme* 
 Lapponici. .hut hietja hahtse ahtse lohlte. 
 
 Syrianic ...hvajt sizim hohjamys ohmys das. 
 Hungarian . hat het nyoltz hilentz iiz. 
 
 Basque. 
 sei zazpi zortzi bederatzi a/mar. 
 
 Caucasian. 
 Georgian . . . ehhvs i svidi rva 
 
 Tuschi yethch wort hart 
 
 Circassian . chi Vie ga 
 Abkhasian ,f bis aa 
 
 Turkish. 
 dlti yedi seldz dohuz on. 
 
 The first point to which I would draw attention 
 here is, the manner in which several of these numerals 
 for " eight" and " nine" are formed. It will be at 
 once apparent, on comparing the Fin, the Esthonian, 
 and the Lapponic expessions for "one" and "two/ 9 
 "nine" and "eight," that in each of the three dialects 
 the last element of " nine" and " eight" is the same, 
 while "one" is the first element of "nine," and "two" 
 of "eight." J Thus, in Esthonian, "one" is ut(s), and 
 
 * Observe, as a basis for inquiry, that the decade comprises 
 only three characters, "one" "five," and "ten." The Fin leak-, 
 the Syrianic kylc, and the Turkish iJci, " two," seem —yk ylc, n. 
 
 f The non-radical Abkhasian suJGx -la is omitted throughout 
 (ante, p. 52). 
 
 J Cf. Pott, Zablmethode, p. 129, note. 
 
Ibl THE ASIATIC AFFINITIES OF 
 
 "nine" is uUtesa; "two" is kat{s), and "eight" is 
 Jcat-tesa. The only solution of this is, that such numerals 
 for "eight" and "nine" are formed on the principle of 
 the Roman nx and ix, duodeviginti and undeviginti ; 
 and consequently that in the Fin -deksa, the Esthonian 
 -tcsa, and the Lapponic -tse, we have three forms of a 
 word allied to the Aryan for " ten," which is, besides, 
 found explicitly in the Syrianic das, "ten 55 (= Ossetic 
 das or dels), and in the Hungarian tiz, "ten." Pre- 
 cisely in the same manner the Syrianic kokjamys, 
 " eight," and okmys, " nine," are formed, and would 
 therefore imply a word rays or amys, " ten," which 
 the Syrianic komyn, " thirty,' 5 neljamyn, " forty, 55 and 
 vitymyn, " fifty," would indicate to exist also under 
 the form myn, amyn, or ymyn. This word seems to 
 me the Fin kymmen and the Esthonian kiimme, "ten," 
 and may be akin to the Basque amdr, " ten," which 
 takes the form ama in ama-ica, " eleven," where led 
 would be "one," and = Sanskrit eka, Abkhasian aka, 
 Hungarian egy, etc. 
 
 The Hungarian and Basque for "eight 55 and "nine" 
 would likewise be compound terms, but of a different 
 nature. As we can hardly avoid connecting nyol- in 
 the Hungarian nyol-tz, " eight," with the Finnish 
 words for "four," such as the Syrianic njolj and the 
 Lapponic nelje, it would follow that nyol-tz, " eight," 
 must = 4 x 2 (compare quatre-vingt), and consequently 
 that ~tz is equivalent to two, zwei, or the Tuschi si, 
 "two." In ki-lentz, the Hungarian for "nine," I should 
 conjecture that lentz is the same as neltz or nyoltz, and 
 that ki-len-tz —1+4x2. 
 
THE OLD ITALIANS. 133 
 
 The next step is to compare the Hungarian and the 
 Basque for " eight" and " nine" : — 
 
 Hungarian. Basque. 
 
 " One/' egy, bat, 1. 
 
 "Eight/' nyol-tz, zor-tzi, 4x2. 
 
 " Nine/' la-len-tz, bed-era-tzi, 1+4x2* 
 
 The Hungarian and the Basque seem here to have 
 the same formation, but to possess only one element 
 in common, which is tz or tzi, " two"; and, as the 
 Basque for " two" is hi, which is to be compared with 
 the Latin bi(s) and the Abkhasian vi, the Basque 
 would apparently contain a complete form for " two" 
 very nearly identical with the German zwei, = Sans- 
 krit dvi or dva. So we have in Hindustani, du, "two," 
 and bet-Teh, " twelve." Again, the Basque forms for 
 " eight" and " nine," if they are composed like the 
 Hungarian, would contain zor and era, "four." These 
 elements, too, may be allied to the Aryan, as " four" 
 is in Sanskrit catvdr or catur, in Afghan zalur, in 
 Hindustani car, in Armenian cor and char, in German 
 vier, and in Swedish fyra. B is the letter which is 
 retained in every Aryan form of " four." 
 
 When we pass from the Pyrenees to the Caucasus, 
 and consider the Georgian for " eight" and " nine," 
 this letter r immediately attracts attention. For, in 
 the Georgian language, "eight" is rva, and "nine" is 
 zkh-ra, which last form = 1 + 8 ; for zJch- would 
 — Tuschi zha, "one," and -ra would = rva, "eight." 
 
 The comparison between the Basque and Caucasian 
 numerals for "one," "eight," and "nine," leads to 
 these results :— 
 
134 THE ASIATIC AFFINITIES OF 
 
 }> 
 
 Circassian ...se 
 
 Lesgi za J-i. 
 
 Tuschi zha 
 
 Georgian r — va "j 
 
 Suanian ar -a > viti. 
 
 Miugrelian ... r — no ) 
 
 Georgian zWi-r — a ^ 
 
 Suanian chh-ar - a > ix. 
 
 Mingrelian ...clxli-or - o ) 
 
 C hat, i. 
 
 Basque < zor-tzi, vin. 
 
 \hed-era-tzi, ix. 
 
 Using the term " Iberian" to include the Georgian, 
 Suanian, and Mingrelian dialects, it may be said with 
 great probability : — 
 
 i. In Iberian and in Basque, as in Hungarian, 
 "eight" = 4 x 2, and "nine" = 1 + 4x2. 
 
 II. Of the three elements, "one," "two," and "four," 
 which compose "eight" and "nine," "one" is different 
 in Iberian and Basque, while " two''' and " four" are 
 the same, and are apparently Aryan as well as Basque 
 and Caucasian : for the Caucasian r-, ar-, and or-, 
 with the Basque zor- and era-, may all be referred to 
 Aryan forms for " four"; and the Caucasian -va, -a, 
 -ao, and -o, with the Basque -tzi, may all be brought 
 out of such Aryan forms as dva, duo, and zwei. In- 
 deed, the Aryan for "two" is explicitly found in 
 Caucasian and Basque ; as " two" is tu in Circassian, 
 si in Tuschi, vi in Abkhasian, and hi in Basque ; so 
 that the Georgian -va, " two," in " eight," is nearly 
 Basque, and the Basque -tzi, " two," in " eight," is 
 nearly Tuschi. 
 
THE OLD ITALIANS. 135 
 
 in. The Hungarian contains the Aryan for "two," 
 under the form -fe in "eight"; and the Hungarian, with 
 other Finnish dialects, contains the Aryan for "one" 
 and " ten" also. 
 
 All this seems as if there were a certain bond of 
 connexion between the three races that preceded the 
 Aryans in the West, the Basques, the Fins, and the 
 Caucasians ; and likewise as if these three races and 
 the Aryans had in very remote ages a common ancestry 
 and a common home.* 
 
 There are other signs in Basque of the use of 
 numerals which are not explicitly found in that lan- 
 guage. Thus "apres demain" is in Basque etzi, which 
 may be akin to the Georgian ze-g, " apres demain," 
 where -g is perhaps to be compared with the Georgian 
 dgJie or the Basque egun, "day." Again, in Georgian, 
 mazeg is " en trois jours"; and, in Basque, etzi-damu 
 is "en trois jours," and etzi-dazu is "en quatre jours." 
 Here the Georgian ma-zeg appears to signify " one 
 after-to-morrow," while the Basque etzi-da-mu and 
 etzi-da-zu appear to signify " after-to-morrow and 
 one," " after-to-morrow and two." If so, then zu, 
 " two," is implicitly contained in Basque ; and mu, 
 "one," and ma, "one," are implicitly contained in 
 Basque and Georgian, and are to be compared with 
 the Armenian mi, mov, me-, " one," and the Etruscan 
 mach,me- y muv-, "one." Da means "and" in Georgian, 
 
 * The Tamil on-badu, ix, is formed like the Fin yh-deksa, "one 
 from ten," out of the Tamil onrru, i, and pattu, x ; as the Tamil 
 aindu, v, and pattu, x, coalesce into aim-badu, l. 
 
136 THE ASIATIC AFFINITIES OF 
 
 as eta } ta, cnda, da, do in Basque. Etzi-da-mu and 
 etzi-da-zu would be formed like the Basque oguei-ta- 
 batj "twenty and one/' and oguei-t(a)-amar, "twenty 
 and ten, thirty." The Georgian for " thirty" is a 
 similar form, ozda athi == ozi-da-athi, " twenty and 
 ten." The difference between the Basque o-guei and 
 the Georgian o-zi, " twenty," is similar to that be- 
 tween the Latin vi-ginti and the Armenian ch-san, 
 where the Latin g is changed for a sibilant. Both 
 -guei and -zi, as well as o-, might be Aryan. 
 
 The following Aryan numerals seem thus to have 
 been detected in Caucasian, Basque, and Finnish : — 
 
 "One" — in Finnish as y\ M-, and egy ; in Basque 
 as -icd ; in Caucasian as aha. 
 
 "One" — in Caucasian as ma-; in Basque as -m?/. 
 
 " Two" — in Caucasian as tu, si, vi, -va, -a, -no, and 
 -0; in Basque as hi, -tzi, and -zu; in Finnish as -tz. 
 
 " Four" — in Caucasian as ar-, -or-, and r- ; in 
 Basque as zor- and -era-. 
 
 " Ten" — in Finnish as tiz, das, -dehsa, -tesa, and 
 -tse ; and possibly in Caucasian as -zi (= tsi), and in 
 Basque as -guei. 
 
 The analysis of numerals is worth prosecuting farther. 
 The most perfect Aryan form for " six" is the Zend 
 hhsvas, otherwise written csvas, which would have 
 passed into Jchvas before it could give the Armenian 
 wez and the Albanian gyas(te). Now Jchvas is like the 
 Georgian elchvsi, "six," but gives no explanation of it. 
 If, however, we interpret ehh-vsi by the Finnish dia- 
 lects, it becomes significant. It would be yh-iviisi, 
 
THE OLD ITALIANS. 137 
 
 
 1 + 5, and = Fin k-uusi, Esthonian k-uus, cc six 
 which, with the three other Finnish forms for tc six 
 may likewise be reduced to 1 + 5, vi. Having got 
 thus far, let us again take up the Zend kh-svas, and 
 suppose it, as well as the Ossetic ach-saz, " six," to be 
 1 + 5. In this case, svas, with the Sanskrit sus and 
 the Afghan sbaz, " six," would properly be " five"; 
 just as the Armenian wez, " six," having lost the pre- 
 fix implying " one," is to be compared with the Fin 
 wiisi, the Lapponic wit, the Turkish bes, and the 
 Basque host, all signifying "five." In like manner, as 
 we have seen (ante, p. 54), the Circassian chi, vi, 
 would be the Etruscan hi and the Lesgi chew a, v; the 
 Etruscan huth, iv, would be the Georgian khuthi and 
 the Lazic kind, v ; and the Abkhasian phsi, iv, would 
 be the Tuschi phchi, v. It seems, then, as if there 
 were once a primeval word, svas, "five," which was 
 common to Aryans and Turanians ; and this word 
 would be found in Basque with its original sense, as 
 the second element of the Basque zaz-pi, vn, would be 
 the Basque hi, n : for zaz-bi would become zazpi, just 
 as ez, " non," and ba, " si," coalesce in Basque into 
 ezpa, "nisi." It is evident, if zazpi be vn, and pi 
 = bi, ii, that zaz must necessarily be v, though this 
 would have been forgotten when terms like the Sans- 
 krit gas were employed for " six." The adoption of a 
 new term, such as pane an, for "five," may have been 
 the cause of such inaccuracy. 
 
 Signs of primeval affinity seem so remarkable here 
 as to deserve being tabulated : — 
 
138 THE ASTATIC AFFINITIES OF 
 
 Hebrew echad, I. 
 
 Hungarian egy } i. 
 
 Sanskrit eka, i. 
 
 Abkhasian aka, i. 
 
 Chinese wit, v. 
 
 Lapponic ] -, 
 
 Esthonian 
 
 f wit, y. 
 
 ut, VI. 
 
 f wiis, v. 
 
 ( h UUSy VI. 
 
 Georgian ekh vsi, vi. 
 
 Armenian g wez, vi. 
 
 Ossetic ach sdz, vi. 
 
 Basque ■< / 
 
 ^ ( zaz pi, vn. 
 
 Hebrew ses, VI. 
 
 Sanskrit sets, vi. 
 
 Zend kh svas, vi. 
 
 Afghan sbaz vi. 
 
 Turkish bes, v. 
 
 Basque host, v. 
 
 Have we any indications of what this supposed 
 primeval word for " five" may have been ? There 
 cannot be much doubt about the most probable mean- 
 ing for such a word. This meaning is " hand": and 
 the apparent affinity between such words as the Per- 
 sian pang, "five," and pane, " fist," has been noticed 
 by several writers. The Basque bost, " five," might 
 thus be related to the Slavonic p>jast, " pugnus," and 
 to the German faust and the English fist ; all which 
 words have nearly the termination of the Zend zasta, 
 
THE OLD ITALIANS. 139 
 
 Sanskrit hast a, " hand." So too the Turkish bes, 
 "five/' which seems akin to the Basque host, "five/' 
 resembles the Gaelic has, bos, "the palm of the hand/' 
 which is the same word as the Welsh bys and the 
 Breton bez, " finger"; terms capable, like the Arme- 
 nian hoyth, " thumb/' = Welsh bawd, of explaining 
 the Basque bat, " one," and perhaps the Turkish Mr, 
 " one." In like manner, we might pass from the 
 Tuschi bid, "fist," to the Chinese wu, "five." It may, 
 too, be possible that both the Turkish bes and the 
 Basque host, "five," are originally allied to the Afghan 
 sbaz, "six" (properly "five"), and to the supposed 
 primitive svas, "hand." At any rate, a word like 
 svas, " hand," seems contained in many languages of 
 different families. It may emerge in the Armenian 
 tluith, " hand, fist"; in the Tuschi tot, " hand"; in the 
 Egyptian toot, "hand"; in the Gaelic dbid, "hand"; 
 and again in the Armenian thiz, " a span," thez-ovk, 
 " a pygmy." It may be seen (ante, p. 105) how sos 
 = thoth in Armenian. Svas may also appear, and in 
 a form more like itself, in the Persian saz, " make"; 
 in the Armenian sos(aphel), "to handle"; in the 
 Phrygian sos(esait), "he makes"; and in the Gaelic 
 sets, "lay hold of." A similar word might be discerned 
 in the Basque escu, the Suanian si, and the Chinese 
 sett, all signifying " hand": and even the Bsthonian 
 hclssi and the Lapponic hat, " hand," the Lesgi koda, 
 "hand," the Ossetic koch, kuch, " hand," leach, " foot" 
 (cf. Armenian fcacA-avel, "to dance," and English 
 hick), and the Tuschi k'hak, " hoof," may bear some 
 
liO THE ASIATIC AFFINITIES OF 
 
 signs of an original likeness to svas, which has become 
 the Welsh chwech, "six, ef." The Lapponic hat, the 
 Lesgi Iwda, and the Ossetic huch, " hand/' would 
 help to explain the Georgian hhuthi, "five," like as 
 the Georgian phehhi, " foot/' is apparently identical 
 with the Tuschi phchi, " five." The Circassian tfchu, 
 "five," and pse, "ten/' seem allied to hhuthi and phchi; 
 and if so, then the Circassian pse would be "feet," as 
 the Abkhasian sva, "ten," might be "hands." Reckon- 
 ing by scores originated, most probably, in men once 
 counting with their feet as well as their hands. Both 
 Caucasians and Basques reckon by scores: thus "forty" 
 is " twice twenty," and so on. The apparent identity 
 of the Abkhasian sva, " ten," with the Basque zaz-, 
 "five," in zazpi, "seven"; and of the Georgian phehhi, 
 " foot," with the Tuschi phchi, " five," the Abkhasian 
 phsi, "four" (properly "five"), and the Circassians^, 
 " ten"; — this shews how " ten" may be the plural of 
 " five," and thus be nearly the same word. We may 
 consequently compare the Tuschi itt, the Lazic wit, 
 the Mingrelian ivithi, and the Georgian athi, all mean- 
 ing " ten," with such Finnish words for " five" as the 
 Lapponic wit and the Hungarian ot. The Finnish 
 words might originally signify "hand" or "foot," and 
 the Caucasian words, "hands" or "feet." 
 
 The five Finnish expressions for " six," kuusi, hints, 
 hut, Icvajt, and hat, are all alike, and all probably 
 = 1 + 5. But, in the expressions for "seven," a 
 difference is discernible. The Lapponic hietja and the 
 Hungarian het } " seven," may = 6 + 1 ; but the Fin 
 
THE OLD ITALIANS. 141 
 
 seitsen, the Esthonian seitse, and the Syrianic sizim, 
 " seven/' seem differently composed, and bear a like- 
 ness to the Georgian and the Basque for " seven": — * 
 
 Fin sei-tse(n). 
 
 Esthonian ...sei-tse. 
 
 Syrianic si-zi (m) . 
 
 Georgian svi-di, 
 
 Mingrelian .. .sqwi-ihi. 
 
 Basque zaz-pi (= zaz-bi, 5 + 2). 
 
 There is no objection to making sei- = si- = svi- 
 = zaz-, as se-decim = sex-decem, and so-dagan = sas- 
 dacan. Similar instances of elision may perhaps be 
 found in the Hebrew sis, se-ba', and se-moneh, " six/' 
 "seven/* and " eight.^f 
 
 Svas, "hand/' especially as we have also the Basque 
 escu and the Suanian si, " hand/' as well as the Ab- 
 khasian sva, "ten," i. e. "fives" or "hands/' will thus 
 bring together the Basque zaz- (which is nearly svas), 
 and the Georgian svi- (which preserves the v of svas, 
 and is like the Abkhasian sva), and the Mingrelian 
 sqwi- (which approaches to the Basque escu), and the 
 Syrianic si- and the Fin sei- (which resemble the 
 Suanian si). In like manner, the Aryan dva, dvi, 
 zwei, bi(s), and 8l($), with the Tuschi si and the 
 Abkhasian v% "two/' will explain the Basque -pi 
 (= hi), and the Mingrelian -thi, and the Georgian -di, 
 
 * It is worth noticing', by the way, that seitse-n and sizi-m have 
 terminations like the Aryan sapta-n and sejpte-m. 
 
 f -moneh seems allied to mdneh, "part, number/' and muneh, 
 "part, time." i( Parts" imply duality at the least. 
 
142 THE ASIATIC AFFINITIES OF 
 
 
 and the Syrianic -zi y and the Fin -tse ; which last two 
 forms would thus = the Hungarian -tz in nyol-tz, 
 4 x 2, and la-len-tz, 1 + 4 x 2, as well as the Basque 
 -tzi in the similarly composed numerals, zor-tzi and 
 bed-era-tzi* The Abkhasian bi~s, " seven/' may con- 
 tain the elements, hi, "two/' and s } "five/' = Suanian 
 si, "hand." In the Georgian, Abkhasian, and Basque, 
 and in the three Finnish dialects, the Syrianic, the 
 Esthonian, and the Fin Proper, there would conse- 
 quently appear to be a similar combination of the 
 same two elements in the number vii; and these 
 elements would belong to the ancestors of the Aryans, 
 as well as to the ancestors of the Fins, the Caucasians, 
 and the Basques. These last three families or nations 
 would, moreover, when they formed their " seven/' 
 have used svas rightly, as " five," not as " six." This 
 cannot be said of the Aryans : for if the Sanskrit 
 sa-pta(n) is connected with sas (which may possibly 
 be the case, though the Zend Misvas and haptan seem 
 
 * The explanation of the Finnish sei-tse, "seven," as 6 -t- 1, 
 from such forms as the Basque sei, " six," and the Circassian se, 
 "one," might be possible, but would hardly be probable. In 
 another Finnish dialect, the Ostiak, ki is n (== Syrianic kyk), 
 vet is v (= Syrianic vit), and ta~vet is vn, i. e. ta + v. There- 
 fore ta is ii, as well as ki, which is used subtractively, as in kyt, 
 in, i. e. ki, ii, from vet, v. Ta~vet, vn, is thus formed out of 
 the same elements as the other Finnish, the Caucasian, and the 
 Basque terms for vn (p. 141) ; elements that are Aryan as well as 
 Turanian. The Ostiak kyda, vi, seems = kyt-da (= ta), 3x2; 
 as nida, vni, would be net-da, 4x2, like the Hungarian nyoltz, 
 the Basque zortzi, and the Georgian rva. The Ostiak net, iv, 
 seems = "one from five (vet)," as the Tamil nungu, iv, seems 
 = " one from five (aindu)." 
 
THE OLD ITALIANS. 143 
 
 against such a supposition)^ yet we could not well get 
 "two" out of -pta* If "two" is found at all in coali- 
 tion with sets, it would rather be in as~ta(n) or as-tau, 
 " eight." As is " six" in Lazic^ and sa is " six" in 
 (Pelasgic) Etruscan. As in the Georgian and Basque 
 pronouns, so too in the formation of the Georgian and 
 Basque for " seven/' an affinity to the language which 
 was becoming Aryan would appear, though the three 
 forms of speech became afterwards very distinct. + 
 It may have been observed that Finnish dialects 
 
 # Compare the Circassian pt-Ve, "four," which seems — " one 
 from five," iv. The Aryan numerals for "three," "four," "seven," 
 and " eight," are not easily explained. " Four" is perhaps the 
 most difficult. 
 
 t Dr. Latham, in his Varieties of Man (p. 127), gives the word 
 khut, "hand," as used in the Manipur and Khoibu languages in 
 Upper Birmah; and he compares it with the Lesgi koda, " hand." 
 It is still nearer to the Georgian khuthi and the Lazic khut, "five." 
 and to the Pelasgic Etruscan huth, " four." Svas, " hand, five," 
 does not appear to be confined to the Old World, for I find in the 
 same work the following Natchez words :— 
 i-spesh-e, "hand." 
 shped-ee, " five " 
 Spesh and shped may = svas, as Armenian spit-zk = Sanskrit 
 (jvet-a, = English whit-e = German weiss ; analogies which shew 
 how the Natchez shpedee, the Lapponic wit, and the Fin wiisi, all 
 meaning " five," may be originally the same word. Compare also 
 ispeshe, "hand," with the Sanskrit spac, "facere," and the Gaelic 
 spdg, "paw." Again, "hand" is shag-Sii in Omahaw, and shak-e in 
 Mohawk ; and " foot" is see and seeh-ah in Sioux, and a-sftoo in 
 Pawnee. Su is "foot" in Chinese. At Norton Sound, near Behring's 
 Straits, "hand" is Sii-shet, "nails" are shet-ooe, and "four" is shet- 
 amik. We may have here, and in the words cited in the text (ante, 
 p. 139), different forms of one of the primitive words of the human 
 race, and a sign of its original unity. 
 
1 I I THE ASIATIC AFFINITIES OF 
 
 ten 
 4chl, 
 
 employ in compositioi] a different word for ' 
 than their own (p. 132). So does the Etruscan in 
 "-genti/' = Ick-lch, 10 x 10. . So does the Lithua- 
 nian in -UJca, "-teen," = Polish lih, "number.** And 
 so, too, does the English in e-leven and twe-lve, the 
 Gothic ainlifsmd twalif. Grimm agrees with Bopp in 
 regarding -lif (and -lika) as forms of a primeval word 
 for "ten/ 1 einer uralten zehnzahl. This word seems to 
 be found, and in our English form -lev en too, in the 
 Malay sa-lapan, "nine/* and du-lapan, "eight**; words 
 which contain the Malay sa, " one** (cf. Circassian se, 
 Lesgi za, Tuschi zlia, "one"), and duiva, " two" (a 
 perfect Aryan form like the Afghan ditva), and are 
 evidently constructed just like the Fin yli-dehsa, "nine,** 
 and kali-dehsa, " eight," ix and nx. As the Malay is 
 thus connected in some points with more western and 
 northern languages, it is possible that it may be so in 
 other points, and thus be allowably employed in the 
 explanation of such languages. Now the Malay for 
 "five** is lima. Prefix to this the hah, " two,** of the 
 Fin hah-dehsa, " eight," = 2 from 10, and we should 
 obtain hali-lima, 2 from 5, "three,** which might be 
 contracted into the Fin kolmi and the Lapponic holm, 
 " three,** nearly as two-leve becomes twelve in English. 
 The Finnish words for " four," such as the Esthonian 
 n-elli, may mean " 1 from 5," jv, and be allied to the 
 Turkish aUi, "six," elli, "fifty,** and el, "hand.** If 
 the Etruscan za-l and the Georgian sa-mi, " three,** 
 are allied to the Fin ho-lmi, and thus imply sa-lmi as 
 a more perfect form, then za and sa would be " two/* 
 
THE OLD ITALIANS. 145 
 
 like the Tuschi si and the German zwei. Compare 
 also the Javanese id-la, "three." At any rate, since 
 there are several ways, as will be more completely 
 shewn directly, of making -I = "five," the Etruscan 
 za- in za-l would most likely be " two," and thus 
 = Georgian sa-; for it seems that the Georgian sa-mi 
 = Mingrelian su-mi == Lazic gu-m = Syrianic hu-jm 
 = Lapponic Ito-lm, " three," i. e. " two from five." 
 And thus the composition and the first elements of 
 the Etruscan za-l and the Georgian sa-mi, "three," 
 would apparently be the same, whatever may be thought 
 of their second elements. If t be "five" in the 
 Tuschi wor-t, "seven," and ba-r-i, "eight," then wor- 
 and -r- would be "two," == Georgian ori, = Chinese 
 ar ; while ba- would be " one," and probably allied to 
 the Basque bat, "one," and possibly to the Hebrew 
 -bo* in se-ba 9 , "seven." As the Tuschi ba-rt, "eight," 
 seems = 1 + wort, " seven," so the Circassian b-gu, 
 " nine," may = 1 + ga, " eight," as the Georgian 
 zlch-ra, "nine," = I + rva, "eight." It is difficult to 
 guess what the Circassian gu or ga, "eight," may 
 have been originally; but, if we were to combine it 
 with the Georgian rva and ra, " eight,'''' we might get 
 g-r, "four," and ua or va, "two." The Circassian 
 would, however, in such a case, want the characteristic 
 letter r of the Aryan " four"; and the Abkhasian a-a, 
 " eight," if = 4 x 2, would have suffered still more 
 than the Circassian g-a. 
 
 It may be as well to tabulate for the second, or I 
 " five," as I have done for the first (ante, p. 138) : — 
 
 L 
 
146 
 
 THE .ASIATIC AFFINITIES OF 
 
 Welsh 
 
 Breton 
 
 Cornish . 
 Armenian 
 
 Turkish . 
 
 Malay .... 
 
 Basque , 
 
 Tuschi . . . 
 Georgian . 
 Chinese 
 Pin 
 
 Syrianic 
 Lazic . . . 
 
 j bys, "finger. 1 
 
 | Haw, " hand/' 
 
 bez, "finger." 
 
 lau, " hand." 
 
 boyth, "thumb." 
 
 el, " hand." 
 
 lima, "hand." 
 lima, v. 
 
 { 
 
 ( hat, 
 
 Mingrelian 
 
 Georgian 
 
 Etruscan 
 
 Javanese 
 
 Circassian pt— 
 
 Esthonian ne— 
 
 Syrianic njo - 
 
 Basque 
 
 Circassian 
 
 $ 
 
 Tuschi , 
 
 ( bar 
 
 i. 
 ii. 
 
 ir. 
 
 ii. 
 
 ii. 
 
 Imi, ii v (ho- 
 
 hi, 
 si, 
 
 ori, 
 dr, 
 ho- 
 hu — jm, nv. 
 
 ffU 7)1, II V. 
 
 su — mi, 
 
 sa — mi, 
 
 za — I, 
 
 tct — In, 
 
 Ve, 
 
 Ill, 
 
 V, 
 
 lau, 
 V le, 
 
 f two"). 
 
 nv. 
 
 IIV. 
 IIV. 
 IIV. 
 IV. 
 
 iv (ne 
 
 IV. 
 
 iv (" one" lost) 
 
 VII. 
 
 
 "one").* 
 
 vx>r- 
 
 
 VII. 
 VIII. 
 
 The Abkhasian bi-s and the Circassian U-le, "seven/ 
 
 * See also note {ante, p. 142), and compare Tamil ndlu, iv (ante, 
 p. 130, note). Nye is (< one" in Albanian. 
 
THE OLD ITALIANS. 147 
 
 appear similar forms, with the same "two/* and a 
 different " five/* Bi-s = 2 + 5, as the Basque zaz-pi 
 and the Georgian svi-di = 5 + 2. The Malay lima, 
 "five/* is evidently the lima, "hand, arm/* of the 
 islets between Timor and Papua.* Like the former 
 word for " five, hand/* this second word seems to stretch 
 across Europe and Asia. For, as the Basque escu- 
 would be the Suanian si and the Chinese seil, all 
 meaning " hand" (as the English shew and shy are the 
 German scheu), so the Malay lima, "hand/' "five," 
 and -lapan, "ten/* are to be compared with the 
 following words cited by Diefenbach (Lex. Comp. s. v. 
 lofa) : Gothic lofa, " the open hand/* = Scotch loof ; 
 Gaelic lamh, Welsh llaiv, " hand**; Cornish lof, lau, 
 "hand**; Gaelic lapadh, "paw"; Polish tapa, "paw**; 
 Lapponic lapa, "the sole of the foot." The same 
 root would also be found in the Welsh Ham, " stride, 
 step/* and in the German lauf and the English leap ; 
 as well as, probably, in the Tuschi lap, " step, stair, 
 treppe" and lam, " mountain.** By a similar associa- 
 tion of ideas, we may connect together the Tuschi it, 
 " run/* and itt, " ten**; and might detect svas in the 
 Hebrew sus, sus, "rejoice, leap" (cf. Polish sus, "hiipfen"), 
 sus, "horse/* and sets, "moth/* = gtjs. Svas might 
 also supply the thema for the Armenian sos- or thoth-, 
 sovth- and hack-, as well as for sos- (p. 105). Of the 
 three characters which compose the decade, i, y, x, 
 
 * Crawfurd's Malay Grammar and Dictionary, vol. i, p. xcviii. 
 Cf. Pott, Zahlmethode, p. 121. 
 
MS 
 
 THE ASIATIC AFFINITIES OF 
 
 the most likely moaning for [ is " finger"; for v, 
 "hand"; and for x, "hands" or "fingers" collectively. 
 The English ten, -teen, as the German zehn and zehen 
 intimate, is " toes/' i. e. " fingers";* and -leven would 
 probably be " hands." The resemblance which the 
 Circassian se, Lesgi za, Tuschi zha, " one/ 5 bear to 
 the German zehe, zeig-, and zeich-, or to the Basque 
 atz, "finger/' should perhaps not be passe#Jtonoticed. 
 The Polynesian dialects are connected wifinxhe Malay 
 family. The following numerals are used in Hawaii 
 and Tahiti f : — 
 
 
 i 
 
 ii 
 
 in \ 
 
 IV 
 
 v. 
 
 Hawaii. 
 
 ..ale alii 
 
 arua 
 
 alcoru 
 
 altaa 
 
 arima. 
 
 Tahiti . 
 
 ..atahi 
 
 arua 
 
 atom 
 
 amaha 
 
 arima. 
 
 
 VI 
 
 VII 
 
 VIII 
 
 IX 
 
 X. 
 
 Hawaii . 
 
 ..aono 
 
 ah it a 
 
 avcmi 
 
 aiva 
 
 urn!. 
 
 Tahiti . 
 
 .aono 
 
 aliitu 
 
 avaru 
 
 aiva 
 
 almru. 
 
 Rima means "hand, arm/' in Hawaii, just as lima 
 does in Malay; and thus explains the Polynesian arima, 
 " five," as the Malay lima, "hand, arm," does the 
 Malay lima, "five." The Polynesian initial a seems 
 superfluous. There is a similar conversion of I into r 
 in the Finnish dialects, where the Esthonian and Lap-' 
 ponic Iwlm, "three," becomes the Hungarian ftdrom, 
 " three," which may enable us to pass to the Basque 
 
 * " Noch unleugbarer stehn BaKTvkos, digitus und zeha (digitus 
 pedis) mit Se/ca, decern, SelKvvfjn und zeigen in zusainmenhang." — 
 Grimm, Geschichte der Deutschen Sproxhe, p. 244. 
 
 f These islands are separated by 2500 miles of sea. 
 
THE OLD ITALIANS. 149 
 
 liiriij cc three. " Compare,, too, the Hindustani so-leh, 
 "sixteen/ 5 and set-re/^ " seventeen."* The Polynesian 
 words for " three" are plainly cc 1 to 2," and not " 2 
 from 5" like the Finnish. The Hawaiian umi, ""ten/ 5 
 resembles the Fin hymmen and the Esthonian Jciimme, 
 " ten/' as well as the Basque amdr or ama, " ten" 
 The Polynesian araa, "two/' is rather like the Georgian 
 ori and the Chinese ar, " two"; and the Polynesian 
 tiva-ru, " eighty" might possibly be compared with 
 the Georgian r-va, " eight/' supposing the two elements 
 reversed, as well as with the kindred Suanian and 
 Mingrelian terms for " eight/' ara and tug. The 
 Lazic ovrOy "eight," is still nearer to avaru. Such re- 
 semblances should be mentioned, though I hold to the 
 former explanation of rva 3 ara, and ruo {ante, p. 134). 
 Avaru seems =4x2, and is thus apparently com- 
 posed like the Hungarian and Basque for " eight/' 
 though with different elements. t 
 
 # L is always represented hj r in Zend. The city of Lima in 
 Peru is so called from the river Rimac. 
 
 f The Malay laid, " man, husband/' may be compared with the 
 following words : — Circassian lay, " flesh' '; lay, t'hlay, " blood"; 
 tlay, "husband"; — Ossetic lag, "man"; lappu, latu, "lad"; — Lesgi 
 les, "man"; less, "husband"; tVyadi, "wife"; — Lycian lade, "wife"; 
 — Abkhasian Ikhadza, " husband"; — Esthonian laps, lats, " child." 
 The Caucasian Leges and Lazi of antiquity were probably the 
 " men." Olalce means " man" in the Arawak dialect of Guiana, 
 which supplies an example of "numeration of the rudest kind," 
 where Jcabo is like the Tamil kai, " hand." See also p. 139. 
 
 Aba da kabo = "once my hand," = five. 
 
 Biama da kabo = " twice my hand," = ten. 
 
 Aba, olake — "one man," = tvjenty. 
 
 (Latfeam, Ethnology of the British Colonies, p. 260). 
 
150 THE ASIATIC AFFINITIES OF 
 
 Witli regard to the symbols, I, v, x, " tlie digit," 
 " tlie hand," and " the hands" or " the fingers/ 5 I 
 would represent accurately enough the extended finger, 
 and v the angular space between the thumb and the 
 forefinger when the hand is held up. This angular 
 space is called in Armenian chil, which is like the 
 Georgian hlieli or qheli, "hand"; both which terms 
 might contain the Etruscan or Pelasgian hi, "five," 
 with the addition in one case of the Etruscan and 
 Armenian termination -il, and in the other of the 
 Georgian termination -eli. x would be the figure 
 formed by placing the two hands across one another. 
 The Chinese character for " ten" is a cross, which is 
 called si, as "hand" is called seu in Chinese, and si in 
 Suanian, and as " ten" is sva in Abkhasian. 
 
 The results of the previous analysis of numerals are 
 to be taken in conjunction with what seems to follow 
 from the numbers on the Etruscan dice, namely, that 
 the Pelasgic Etruscan numerals were Caucasian. See 
 especially the tabular view (ante, p. 54). The infer- 
 ences which I should be inclined to draw from the 
 numerals, as well as from conjugations and pronouns, 
 have been already explained in my first chapter, 
 where I have brought together several other coinci- 
 dences of different kinds, which appear by their com- 
 bined force to conduct us to a similar conclusion. And 
 this conclusion would be : — that before the Aryans 
 began to spread from their original home, they dwelt 
 there with Fins and Caucasians on their west ; the 
 
THE OLD ITALIANS. 151 
 
 Caucasians tending towards the south, and the Fins 
 towards the north : and that as the Fins scattered 
 themselves, speaking in a general manner, over the 
 northern half of Europe, the Caucasians did the same 
 over the southern half, but probably at an earlier 
 period ; for the Caucasians, especially if it were allow- 
 able to include the Basques among them, cannot be 
 said to have developed a common numeral system 
 before dispersion, while the Fins would have done so, 
 though not quite as perfectly as the Aryans. Both of 
 the Turanian races would have been continually im- 
 pelled farther westward, as the Dravidas would have 
 been southward, by the expansion of the Aryans, who 
 ultimately broke through the Western Turanians by 
 two different routes, one on each side of the Buxine, 
 and gradually encroached upon them till they were 
 left as they now are, in the Caucasus, and the Pyrenees, 
 and the North of Sweden and Russia, though their an- 
 cient presence in the heart of Europe is still indicated 
 by two or three words used in the Alps. When Livy 
 attributed Etruscan affinities to the Alpine population 
 in general, but especially to the Ehsetians, he probably 
 spoke with more accuracy than has been generally 
 thought, or even perhaps than he himself was aware 
 of. For all, or nearly all, the original inhabitants of 
 the Alps (as well as of the pile-dwellings on the 
 Swiss lakes) may have been Tuscan,\.e. Caucasian, while 
 the Aryan Rasence penetrated no farther to the west 
 than Khgetia, and a subsequent Celtic inroad made the 
 Aryan population of Noricum quite as much Celtic as 
 
152 ASIATIC AFFINITIES OF OLD ITALIANS. 
 
 Thracian. In Armenia and the Caucasus, Asia may 
 thus claim both elements of the Etruscan people as her 
 own, whether they were of Tyrrhenian or of Pelasgian 
 origin. Such, at least, is the hypothesis which seems 
 to explain all the evidence that I have brought forward, 
 and to solve at the same time four ethnological pro- 
 blems. In ancient ethnology, we are led to ask, who 
 were the Etruscans, and who were the Pelasgians ? 
 and, in modern ethnologj^ w T hat has become of the 
 two races of which the Armenians and the Caucasians 
 are the surviving representatives ? Each pair of ques- 
 tions supplies the answer to the other pair. 
 
INDEX OF ETRUSCAN WORDS. 
 
 AcJirum, " Hades, Acheron/' 87. 
 
 Alpan, " supplex," 78, 91. 
 
 Am, " sura," 100 : or see s. v. mar. 
 
 (Ap)avenke } " Aeponit, relinquit," 38. 
 
 Ami, " setas," 27, 28, 30. 
 
 Avils, " setatis," 27, 33, 37. 
 
 Chiem, " quinque," 49. 
 
 Chimths, " quindecim" (?), 49. 
 
 Chiseliks, " monumeiijmm, ^W^ara," 79, 92. 
 
 Eka, « hie, ecce," 6lM? y 90. 
 
 Epana, " epuluin/'TuB. 
 
 Erai, " hilaritatis," or "hilaris," 103. 
 
 Erske, "sese offert," 87,* 
 
 Etera, " alter/' 59. JT 
 
 Ethe, " si, quando,"U02. 
 
 Fleres, "oblatio, dojfcim," 79, 80, 86. 
 
 Flezrl (qu. flerzl), " fblatum, datura/' 86. 
 
 Fuius, " vl6s," 60. 
 
 Helephu, " effundit," or " effunditur," 105. 
 
 Huth, " quatuor," 51, 54. 
 
 Kana, " siraulacrura," or "statua," 83. 
 
 Karutezan, " quatuordecira" (?), 49. 
 
 Kealchls, " quingentos," or " quingentesimi" (gen.), 41, 43. 
 
 Kecha, "expiat," or " solvit," 78, 80, 90. 
 
 Ken, « ut," 79. 
 
 Kepen, "tumulum," 36 (note). 
 
 Kerinu, " sculpit," 90. 
 
 Kethu, " aquse" (gen.), or " aquam/' 97, 107. 
 
 Ki, " quinque," 51, 54. 
 
 Kiemzathrms, t( quinquagesimi tertii," 39. 
 
 Kis, " P€Kp6s," — or else " menses," or " raensis" (gen.), 47. i- 
 
 Kisum, " vatpov," 47. 
 
 M 
 
1 5 1 INDEX OF ETRUSCAN WORDS. 
 
 S >.', u vtKpoTs" or " veicpSs," or " moritur," 47. 
 Klalum, " raoeroreni, funera," 37 (note). 
 Klan, " soboles," or ' ' princeps," 59. 
 
 K l l2si}''V inS > rite '" 78 ' 79 > S0 > 89 ' 90 * 
 
 Kver, " soror," 88. 
 
 Leine, " vivit, fit/' 27, 28, 29. 
 
 Lenache, " facessit, fieri facit," 91, 92. 
 
 Line, " vivebat, vixit," 29, 69. 
 
 Lisiai, "linguae" (dat.), 100. 
 
 Lthas, " Aitt)s" 78, 80, 91. 
 
 Lupum, " cadaver, corpus," 37. 
 
 Lusni, " luminis," 94. 
 
 Ma, " sed," 99. 
 
 Mach, "unus, 51. 
 
 Machs, "mensis" (gen.), or "menses," 41, 44. 
 
 Mar, " vas, fass," 100 : or else — . 
 
 Mar am, " contineo, ich fasse," 107. 
 
 Mathu, " vini" (gen.), or " vinum," 99, 107. 
 
 MuvtuL }" centum/' or "centesimi" (gen.), 41,42,43. 
 
 Mi, " ego, me," 60, 80, 97, 99, 103. 
 
 Nak, " ad, nach," 87. 
 
 Nastav, " hospes," or " hospite," 104, 108. 
 
 Nesl, "mortuus," 61. 
 
 Nethu, "liquoris," or "liquoretn," 103, 104, 107. 
 
 Ni, " non," 97. 
 
 Puia, " filia, evyaTrjp," 60. 
 
 Puiak, "figliuola, Qvyarpiov, tochterlein," 60. 
 
 Puiam, " filiam," 60. 
 
 Puil, " tckpou" 60. 
 
 Rasne, " ulna," 49. 
 
 Pal, " annus," 27, 28, 30. 
 
 Sa, "sex," 51. 
 
 Sak ) 
 
 Sech [ "proles," 60. 
 
 Sek ) 
 
 Salthn, " fusio, fusum, Tdpeupa, opus," 94. 
 
 Sansl, " libens," 79, 88. 
 
 Sas, " sex," 38. 
 
 Semphalchls, "septingentos," or " septingesimi" (gen.), 41, 43. 
 
 Sie, " sit," 103. 
 
 
INDEX OF ETRUSCAN WORDS. 155 
 
 Suthi, "conditur," 61. 
 
 Suthilc I " se P u l crum > tumulus/' 69, 90. 
 
 Suthina, " dvcria" 83, 84. 
 
 Teke, "facit," or « ponit," 79, 91. 
 
 T6n^}" tenet ' tendit > fert * offert," 79, 90. 
 Tesne, " decern/' 49. 
 Tesnsteis, " centum/' 49. 
 Thapna, " koiWt??, lampas," 94. 
 Thipurenai, "calidse, sitienti," 101. 
 Thu, "duo/' 51. 
 
 Thi} e<,r{,iros > si g Mm >" ™> 80, 91. 
 T/mi, " memoratur," 69. 
 Thunesi, " diei/' 42, 45. 
 Tinskvil, " Jovi(s) donum/' 93. 
 Tiers or £wrs, "triginta," 38, 39. 
 
 TZe^asie? S }" debitum P retium > meritum," 78, 86. 
 Tular, " sepulcrum," 68. 
 T:«L}" dat '"80,87. 
 
 Tuthines, " gratise, x^P lT0S >" 78, 79, 83, 84. 
 
 Fta, "filia/' 60. 
 
 £aZ, " tres," 51, 55. 
 
 Zelc, " statua/' 88 : or perhaps zeke, " affert." 
 
 lESej "frfcator, sepelitur/' 34 36, 47. 
 
 THE END. 
 
 T. RTCHARDS, 37, GREAT QUEEN STREET. 
 
ft 8 1 
 
 % 
 
 

 
 
 
 
 
 
 

 
 «* 
 
 
 
 
 
 £ 
 
 
 % .«*' 
 
 
 ^ <<, 
 
 
 
 1 ^ 
 
 *V 
 
 
 
 
 
 
 
 
 ^ 
 
 
 ^ ^ 
 
 1 . v