^{page 229 note 1} The intelligence that admits composition in άνδρό-μεος (Brugmann, , Gr. ii. 1, p. 13)Google Scholar and rejects it in Lat. diu-tinus (ib. § 197: Kvg. § 382, however profound, is inconsistent to the point of irrationality.

^{page 230 note 1} Mere inertia prevents the rectification of many linguistic dogmas. Recently, however, improvement has been made by analyzing άγχι-στîνος as a compound (cf. AJPh. 37, 65²), instead of as a derivative of άγχι-στον, the latter itself a compound = ‘prope-stans’ (see AJPh. 33, 392; 34. 15). The derivation of IE. superlatives in - istho- from is [reduced grade of compv. in (i) yes] + to is the vaguest of glottogonic devices. Some years ago, when I explained the type of Skr. mάmh-i-sthas as ‘in-dando-stans,’ I failed to discover its synonym mamhane-sthā-s ‘liberalis-simus,’ for which Grassmann long before had presented precisely the same analysis. Likewise Sāyana defined νακsane-sthā-s in RV. by ‘vahnau (loc.) sthitah,’ i.e. precisely by ‘in-sacruficando-stans’ (note νāhas ‘darbringung’; and interpret νακsαņe-sthās by the synonymous Agni epithet of haνir-ν̅t, quasi ‘libationem uehens’). And certainly karmani-sthā-s means ‘in re diuina stans.’ The actuality of i in these infinitival and adverbial priora; of the aspiration in Skr. -i- sth-(cf. on λοîσδος, AJPh. 37, 68); of the formation in -sthā, parallel to stha-; of the ancient definition of Sāyana, possibly backed by tradition and certainly informed with a feeling for his own tongue and its literature: all this evidence of fact is to be ignored because of inert acquiescence in a hypothetical conglomerate suffix is + to.

^{page 230 note 2} The copula esti was a mere demonstrative, and meant ‘here by him’ (cf. Chinook paradigms in , Boas, Handbook of American Indian Languages, p. 618Google Scholar; reprinted in Bull. Univ. of Texas, No. 263, § 47). In fact, IE. esmi ‘sum’ is identical with the IE. locative esmi ‘here’ (egodeictic; cf. Ital. ecco mi, and see Bull. § 46.