It was almost certainly this wider notion that was foremost in the minds of those who wrote that inscription and made the dedications which it commemorates, as is abundantly clear from a perusal of the inscription as a whole, and we may assume that the joint temple shared by the Nolans and Abellans was similarly frequented more by farmers and small traders than by triumphing generals. The tithes offered there would belong mainly to the classes (1), (2) and (4) enumerated above, and the official appointed (by the Nolans) to receive them and (perhaps) to administer the proceeds was naturally called deketasiúi.

HITTITE h INITIAL = INDO-EUROPEAN bh

EDGAR H. STURTEVANT

YALE UNIVERSITY

The relationship of Hittite and Indo-European has been recognized chiefly on the basis of certain correspondences in inflectional terminations, suffixes, and pronouns. Indo-European etymologies have, to be sure, been suggested for a considerable number of Hittite words, but a large proportion of these etymologies are clearly unsound and even those which are most plausible have never been worked into the sort of consistent system which could bring conviction to those who understand sound etymological method.

Nevertheless our knowledge of the Hittite vocabulary has reached a point where it ought to be possible to discover Indo-European etymologies even if they are obscured by phonetic change. In this paper I hope to connect several words with their Indo-European etymons, and to establish at least one phonetic law.

The Hittite verb which appears sometimes with the stem huwa- and sometimes with the stem hui- shows a rather bewildering range of meaning. The first to be established was the meaning 'flee', as in the phrase in the law code, tak-ku IR-iš hu-u-wa-i 'if a slave runs away'.

The causative huinu- seems to carry the same force in the annals of the first ten years of the reign of Mursilis II = Keilschrifttexte aus Boghazköi 3. 4. 2. 69 = Hrozný, Boghazköi-Studien 3. 194. The text is incomplete at this point, but it seems fairly clear that when Dapalazunawalis fled (kattan huwaiš) from Puranda he 'caused his [infantry] and cavalry to flee before him' (piran huinut).

· The precise phonetic value of such forms as hu-u-i-ya-mi, hu-u-i-er, etc. must for the present be left undetermined, and so must their relationship to forms from the stem in a (hu-u-wa-a-iš, etc.; cf. Friedrich, Staatsverträge des Hatti-Reiches in hethitischer Sprache 31°, 914). The Hittite vowel system will probably cause more dfficulty than the consonants; we do not know, for example, whether Hittite separated from the parent stock before or after the changes which gave rise to Ablaut. In this paper I shall avoid discussing vocalism.

· See Hrozný, Die Sprache der Hethiter 30, Code Hittite 18.

3 In this paper Sumerian ideograms are printed in capitals, and all Assyrian words or syllables in Italic capitals.

While the phrase piran huwa- (without causative suffix) is very common, it seems never to mean 'flee before', but always 'march before'. It is used, for example, of a marshal conducting high officials to their positions during a religious ceremony. Friedrich, Aus dem Hethitischen Schrifttum 2. 5. 36, 8. 20, 25, etc., translates the phrase 'voranlaufen', no doubt because he regards 'run' as the primary meaning of the verb; but 'march' would evidently fit the context better. That is the meaning required also in the military use of the word, as in the fuller annals of Mursilis (3. 21), according to Forrer's (Forschungen 1. 57) restoration. The king says: 'Since it was not possible to go up <the mountain> with horses, I marched before (piran huiyanun) the army on foot'. Usually the phrase has a god or several gods as subject, and the context makes it perfectly clear that in marching before the army of the Hittite king the gods insure his victory. This interpretation gains support from many Assyrian passages which represent the gods as marching before or beside their royal worshippers.

Difficult as it is to combine in one verb the meanings 'flee' and 'march <to victory>', the evidence seems to require that we do so. Perhaps the original meaning was 'run' or 'hasten'; but it is equally possible that we should start with the meaning 'flee'. For the change of meaning, compare Latin fugio, as in Statius Thebais 9. 770: nullum sine numine fugit missile, and especially Seneca Epistulae 108. 25: Numquam Vergilius . . . dies dicit ire, sed fugere, quod currendi genus concitatissimum est. The further development of meaning from 'run' to 'march' is surprising; but from that there seems to be no escape. The word has other meanings, however, which can scarcely be derived either from 'run' or from 'flee'. One of the curses in the military oath runs: 'May his wives not bear sons and daughters; on the plain and in the... and in the meadows(?) ú-el-ku(?)-wa li-e hu-wa-a-i for him, and may his cows and sheep not bear calves and lambs'. Friedrich shows that the second of the three clauses should be in effect 'may not crops grow for him', and that welkuwan actually appears as the name of a plant <I should rather assume the meaning 'vegetation'> in a parallel curse (4. 17): ú-el-ku-wa-an li-e ú-iz-zi 'may welkuwan <i.e. vegetation> not come up'. Nevertheless Friedrich does not venture to translate huwai as 'grow'," since he can cite no parallels

See Hrozný, BoSt. 3. 17511; Sommer and Ehelolf, BoSt. 10. 81.

KBo. 6. 34. 2. 38-41.

• Zeitschrift für Assyriologie, Neue Folge 1. 181f. (1924).

1ZA, NF. 1. 165 (1924), Aus dem Hethitischen Schrifthum 2. 18 (= Der Alte Orient 25. 2. 18).

=

for such a meaning. He had overlooked the vocabulary entry3 GÚ.ME. [IR.ME.IR HJA.NA.BU = hu-wa-li-ya-[war]. HANĀBU is defined by Muss-Arnolt, Assyrisch-Englisch-Deutsches Handwörterbuch 325, as 'sprout, grow luxuriantly, abundantly'. The verb huwaliyacontains the same formative elements, for example, as irmaliya- ‘be ill', which is a denominative in iya from irmalaš 'ill', while the base of irmalaš appears in the adjective irmas 'ill'. The fact that irmaš is an adjective does not make it necessary to assume an adjective *huwaš 'growing'; for there is no lack of derivatives in 1 from verb-stems. Examples are: arkammanalliš 'tributary'10 from the verb implied by arkammanatar 'payment of tribute';" išhiul 'obligation, contract' from išhiya- 'bind';12 takšul 'peace, peaceful'13 from takš- 'join"; waštul 'harm, injury' from wast- 'do harm'.15 We may then confidently translate welkuwa le huwai 'may plants not grow'. For the present we can scarcely decide whether welkuwa is a nominative singular neuter, equivalent to welkuwan, or whether it is a nominative plural with a singular verb, as in Greek, 16

.12

Furthermore, the participle huwanza appears as final member of compounds in the sense of 'growing, becoming.' Friedrich has shown that the Hittite word for old man is miyahuwanza. The prior member of this compound is the stem miya-, which Friedrich18 understands to mean 'wachsen, blühen'. In all the passages cited the meaning 'be ripe' would fit as well, and in several it is better. Particularly in Keilschrifturkunden aus Boghazköi 8. 35,1o the meaning 'be ripe' seems necessary. The text records the significance of birth in each of the several months, and each paragraph begins with a clause in the follow

19

ing form; i-na ITU IX.KAM DUMU-áš mi-ya-ri 'a child is born in the ninth month'. As Friedrich correctly observes, the meaning 'be born' is the only one possible; but such a meaning can scarcely be derived from 'grow'. The embryo grows before birth, and the child after birth; growth is actually checked for a time at birth. Clearly the word means 'be ripe' and hence 'be ready for delivery'. Elsewhere the participle is used of a vineyard full of ripe fruit, and the iterative-intensive miyešk- means 'ripen'. The derivative miyatar means 'ripe fruit, crop'. The compound miyahuwanza, then, originally meant 'growing ripe' and then 'aged'.

Another compound in -huwanza is šarhuwanza 'womb, embryo'.20 Its prior member is the adverb šar(a) 'up', so that the word properly denotes 'that which grows big'. The final huwanza 'becoming, growing' in these two words may be a participle from the stem huwa-, in spite of the fact that we have a different participle in the phrase piran huiyanza 'helper'.21 We do not yet understand the relationship of the two stems huwa and hui- (cf. fn. 1).

Furthermore our word, in the form hui-, means 'live', in the derivative noun huitar 'the living things, the animals'. The suffix tar usually forms abstract nouns, and so it is probable that the collective force comes from an earlier meaning 'life'.

This derivative forms a connecting link between huwa-, hui- and the stem huis- 'live'.22 The additional element in huis is the suffix (e)š, (2), which appears to strengthen verb-stems in several cases without greatly changing their meaning.23 Thus we find halziš(š)-, halzeš(š)beside halza- call',24 and peššiandu beside pa-, pi- 'give'.25 From au'see'26 we find forms in & in the third person singular: pres. aušzi, pret. aušta, imperat. aušdu, and also in the first person preterit middle aušhahat. From ki- 'lie, be placed' there is a derivative kiš(a)- 'become, be'. The verb hatk- 'shut',27 forms the causative hatkešnu-. As we shall see presently (f. n. 31), this suffix (i) is an element of the iterative-intensive suffix (i) šk.

20 Friedrich, ZA NF. 1. 185 (1924), Zimmern, Orientalistische Literaturzeitung 25. 298 (1922).

21 Hattusilis 2. 39, p. 18 Götze.

22 See Friedrich, OLZ. 26. 46-9 (1923).

23 See Sommer, BoSt. 7. 22.

24 See Hrozný, BoSt. 3. 863; Sommer and Ehelolf, BoSt. 10. 58.

25 See Sommer, BoSt. 4. 151 and especially 7. 22.

26 See Forrer, Zeitschrift der Deutschen Morgenländischen Gesellschaft, NF. 1. 214 (1922); Friedrich, ZA, NF. 3. 1861, 202 f. (1926).

27 See Sommer, BoSt. 7. 1ff.