oris moris followed the analogy of flōs floris (v. § 42, p. 58 supr.) labos laboris etc. and these with the words. which became monosyllables on contraction (rus thus aes plus crus (?)) together with those in which the r was regular as preceding the i of the stem and following an i or u or long vowel (glis glirium etc. mus muris (Skt. musika) jus juris) established the rule for monosyllables whose nominative ended in s, unless there were some neighbouring word to protect them (as cos, cotis, cautes). The form speres quoted from Ennius is an example of this influence, further helped by speráre. heri is the locative (cf. temperi) of an old Latin* hes. Lares Lases: Lar: *las exactly as mares : : mas and arborem arbosem arbor arbos, the old nominative las being lost the sooner because the word was generally used in the plural. dare may have been influenced by the compounds pródere dédere cet. as well as by the regular -re. fore forent have been discussed § 26, p. 33 n. supr. 4. List of words in which the change appears to be due to i and u. 59. a. Words where s has become r after an accented syllable between i and i: With equiria contrast equíso. For gerit etc. v. supr. § 42, p. 57. They would be also influenced by cóngerit etc. The second i of dirimit dirhibet stands on a level with that of rõsidus luridus § 54, p. 75 supr. The words would be also influenced by dirimebat dirhibere cet. Papirius. As to the accent of proper names v. § 46, p. 64 supr. For the exceptions Calvisius etc. (Sabine names) v. § 34, p. 48 supr. B. Where s has become r after an accented syllable between u and u. núrus gen. nurūs. Gr. vvós, Skt. snușa. y. Where s has become r after an accented syllable between u and i. júri etc. tellúri etc. múri etc. haúrit etc. (e)úrit etc. luridus v. supr. § 54, p. 75. Curio Furius Spurius (?) v. § 51, p. 70 supr. haurit urit would be also influenced by urébat etc. 60. S. Words in which s has become after a long accented syllable before i or u. and some others where the long vowel is or u, which have already appeared under (a) or (y). caerulus is the only word which needs comment. Brugmann derives it from caelum by dissimilation of the first 1, which is of course possible enough, and we may in that case dismiss the word. But the constant conjunction 'caerula caeli'2 would be rather meaningless, and the parallelism of ruber 'red', rufus 'red-haired', caerulus 'blue', caesius 'blue-eyed' (caesia 1 The first syllable of pāricida can hardly be called unaccented in the same sense as that of sororem. In such polysyllables we have evidence in Romance that a bye accent was kept on the first syllable. 2 If in spite of this there is a connexion between the words it seems more likely to be that caelum is for * caeslum. virgo = 'Minerva' ('a Sabinis')) seems to connect it with the clearly borrowed word caesius. On the latest system of accentuation, which was probably fairly established by 200 B.C., the accent on caeruleus (which we find in Ennius side by side. with caerulus) would be on the third syllable from the end and this would therefore have been preserved from sinking to u if it had been an original o, since the change from u to o before in unaccented syllables took place quite late, as is evidenced by the frequent variation in spelling, e.g. Aesola Aesula, and therefore after the final change of accent. It may be thought that the u of caérulus had influenced caeruleus, but the influence is more likely to have been the other way. It is usual to find the older form preserved in such cases. In caesi-us caeru-lus may we not see an example of the common parallelism of and u stems in the same or different languages? The stems in Latin have all disappeared in various ways (e.g. suūvis vacuus), but traces of them are left in words in -lo- and -ro- e.g. anguis: angu-lus, inula: Castrum Inui (cf. Mapaðir), oculus: wкvs, ungula: ungu-is vidulus: viduus, also ancora (orig. *ancura; does the change of u to ŏ before r point to an original s, *ancus, * ancoris? v. supr. § 42, p. 57): ἀγκῦρα (for *ἀγκύρια). For the formation cf. ἀγκύλος Sa(o)vλos κаμmúλos. The u stems like tribus (tribulis) are of course distinct. There are a few words in Festus about which nothing is known and which may very well be Sabine like fasena: murgisonem 'a mora et decisione' (!) (perhaps to be placed with equiso supr. § 54, p. 74) adasia 'ovis vetula recentis partus'. C. Evidence as to the change of accent. 61. Since all the words in this list have occurred before and have been fully discussed it seems unnecessary to add references, since they can be at once ascertained on reference to the index accompanying the essay, which has been con 1 v. supr. § 44, p. 60. structed for me with great kindness by one or two friends. A query denotes that the word had probably not reached the form in which it appears in the list (in other respects than the change of s to r) before the beginning of the period of rhotacism, eg. pacari? which before 450 B.C. was probably *pacáiěsi. For the method of arrangement v. § 47 supr. pp. 65-68. 1. Words in which s which became r follows a syllable which was unaccented only on the oldest system of accentuation. lábōris? etc. pléōris? etc. pácari? etc. télluris? etc. ménsarum etc. The proper names Rabirius? Papirius? Etruria? Pinārius? Fálērii etc., stand on a different footing. 2. Words explicable either by the oldest or by the intermediate but not by the latest system of accentuation. Words with r. a. *glé(u)oria; Valerius Véturius cet. 3. Words explicable either by the oldest or by the intermediate or by the latest system. An accent in () denotes one probably not prevailing at the time of rhotacism. |