342 



DISCOVERY 



best of all records — the task is comparatively simple. 

 Between Latin, the language of the earlier population 

 at Rome, and Osco-Umbrian, as the language of the 

 other group is called, there are clearly marked differ- 

 ences. The most striking is the treatment of one 

 class of the original Indo-European consonants, those 

 technically known as the " velar plosives " (like 

 English qu in " queen " and gw in " Gwatkin "). 

 In Latin these sounds were in general preserved, with 

 a following w-sound ; in the Osco-Umbrian dialects 

 they became labials, p and h. For example, Latin 

 quis, " anyone," corresponds to Oscan pis ; so also 

 Latin quanta, "how large," to Umbrian />a«ia ; Latin 

 quinqiie, " five," to Osco-Umbrian pompe. The same 

 difference separates the two subdivisions of the 

 Keltic group of languages (Irish coic, but Welsh pump, 

 " five "). This separation of the dialects into p- 

 dialects and 9-dialects on the basis of the classification 

 just described is excellent so far as it goes ; but it 

 fails us in those cases where the records are too scanty 

 to give us even a single certain example of p or qtt 

 arising from an original q. In these cases we have to 

 seek other evidence of the character of the language 

 of the people into whose antecedents we are trying to 

 inquire. 



Now it has been very acutely obser^'ed that the 

 ethnical or tribal names of the two strata of population 

 are marked by the use of particular suffixes. The 

 names of tribes which were members of, or had been 

 absorbed by, the incoming northern people are dis- 

 tinguished by the suffix -NO- (nom. pi. -ni) as Sahi-ni, 

 Luca-ni, Hirpi-ni, Roma-ni, Campa-ni, Igiivi-ni, and 

 many others. This suffix is found all over Italy, at 

 all events the peninsular part of it ; but its frequency 

 varies considerably in different districts. Amongst 

 the Hirpini it is by far the most frequent ; in Campania, 

 Lucania, Latium, and amongst the Sabines it is 

 preponderant, whereas in Umbria it is less common 

 than its rivals. The earlier stratum, on the other hand, 

 is to be connected with the suffi.xcs -co- and -(a)ti-, 

 for example, Falis-ci, Aiintn-ci, Manu-ci {total tnar- 

 oucai, " for the Marrucan people," on an inscription of 

 the fourth century B.C.) and Ardea-les (the people of 

 Ardea), Tibur-tes (the people of Tibur, the modem 

 Tivoli), and neuter words like Rca-te or Prccnes-te, and 

 in Rome itself Qiiiri-tcs. Both of these suffixes also 

 appear amongst the peoples of Northern Italy, where 

 the population is known from archaeological e\ndence 

 to have been closely related to the early inhabitants 

 of Latium. It is to be observed, further, that this 

 group of names, which appears in greatest frequency 

 on the west coast and in the centre of Italy, often 

 suffered a later extension by the addition of the -NO- 

 suffix, as the peoples bearing them gradually came 

 under the dominion of the Romans (Roma-ni) or. 



earher, of the Samnites (who called themselves Safi-ni, 

 " Safincs "). Thus the Marru-ci became Marru-ci-ni ; 

 the people of Ardea are no longer Ardea-les, but Ardea- 

 ti-ni, and of Ari-ci-a Ari-ci-ni. In Teate, in Apulia, 

 -Ti- gave place altogether to -no- and Teate became 

 Tea-niim.^ The same process is extended even to non- 

 Italic communities, and the citizens of Naples, who in 

 Greek called themselves NeapolilcB, were called by the 

 Romans Ncapolitani. The old ethnicon of the people 

 of Latium, Latiaris (as in Latiaris coUis) is dropped 

 for Latini after the Sabines have made themselves 

 masters of the land, just as in Rome itself we have 

 Quiri-tes and Roma-nus in the phrase popidus Romanus 

 Quiritium. The reverse process, too, may be seen at 

 work in a case like that of Iguvium in Umbria, whose 

 people called themselves first Iguvi-ni and then Iguvi- 

 n-ates, presumably because the -no- invaders had been 

 at some time replaced in authority by a risorginxenio, 

 as it would be called in modem Italy, of the people 

 they had once conquered. 



In a few places in Italy, such as Bologna, where 

 archteological deposits are found one above another 

 in layers representing successive ep>ochs of time from 

 the early Bronze Age down to the Roman period, after 

 the remains of the Roman, Gallic, and Etrascan strata 

 have been removed, there are found two strata which 

 have been recognised as resf)ectively the remains of 

 the two stocks of population we have been discussing. 

 Though there are many things which still remain 

 obscure from the arch.tological standjxjint, this much 

 may be taken as certain, that the lower- of these layers 

 represents the Bronze and preceding Ages, the upper 

 the Early Iron Age. It is clear, too, that the northern 

 invading race brought into Italy the use of iron, and 

 that in the north of Italy they were on the soil before 

 the Etruscans. Thus in the lowest deposit we find 

 that cutting implements are all made of stone or 

 bronze, while iron is unknown ; in the next stratum 

 above it weapons made of iron, axe-heads and the 

 like, are beginning to appear with considerable fre- 

 quency. Above this comes the Etruscan stratum, 

 then the Gallic, then the Roman. 



We must now return to Rome. Livy relates with 

 grave humour the famous and popular legend how 

 King Romulus obtained Sabine women as wives for 

 his men bj' inviting a host of Sabines with their famihes 



' Livy, not knowing this, actually in one place relates the 

 same event twice over in the same chapter (ix. 20), because 

 some of his authorities spoke of Teales (§ 7), the others of 

 Teaiie»ses (54). — K. S. C. 



^ Contemporaneous with the lake-dwellings and terremare 

 described by Professor Halliday in the August number of 

 Discovery, pp. 236 ff., and characterised by objects mostly of 

 the terramara types, although at Bologna not an actual 

 terramara, but a group of hut-villages, has been found. The 

 significance of the different form of dwelling is not clear. 



