1880.] A. F. R. Hoernlc — A Collection of IlimU roots. 



41 



The other is the still more common root dv (or H. H. a) "come." 

 Its origin has, I think, not yet been satisfactorily explained. One would 

 naturally connect it with the Skr. root d-ycl, from which, clearly the 

 Marathi root ye " come" is derived. But this does not explain the termi- 

 nal consonant v in the Hindi dv. Now it is a curious fact, that the root 

 dv imitates, in every respect, the conjugational forms of the root pav (Skr. 

 pi'dp = pra-dp) , instead of those of the rootja "go" (Skr. yd) which one 

 would expect it to follow. Thus, present participle E. H. dvat or W. H. 

 dvatii "coming," E. H. pdvaf or W. H. pdvatu* "obtaining," but E. H. 

 Jdt or W. H. jdiu " going ;" past participle E. H. d'il or dyal or dvd, 

 W. H. dyati " come," E. H. pail or pdyal or pdvd, W. H. pdyau " ob- 

 tained," but E. H. ydll or y^yal or gnyd, W. H. yayau " gone ;" 3rd sing, 

 present H. dvai, H. H. dve "he comes," H. pdvai, H. H. pdve "he 

 obtains," but YL.jdy, H. H.jde " he goes." I incline, therefore, to think 

 that thei'e is here another instance of the, already noticed, tendency of 

 the Indian Vernaculars to assimilate verbal forms, so that the v in dv is 

 due to the influence of ^Jrt'y ; an influence, natural enough, when it is 

 remembered that v, equally with y, is often inserted between two adjacent 

 vowels for the sake of euphony.f This assimilation is a very old one. 

 There are traces of it in Prakrit as well as in the Gipsy dialects. In Pra- 

 krit there is the 3rd sing. pres. dvei,% and shortened dvai (H. C. 4, 367) 

 " he comes." The regular Prakrit form would be ddai or shortened dd'i 

 (see H. C. 4, 240) ; but just as there is uttJie'i or shortened utthal (H. C. 

 4, 17) for uttJidai or uttJidi (see Vr. 8, 25) " he stands up" (of root ut-sthd), 

 so there might be de'i or dai (of root d-yd), from which, by the insertion of 

 the connecting consonant v, there would arise dve'i and dvai.§ 



The following List of Plindi Eoots is arranged alphabetically, in two 

 parts. Part I contains primary roots, while Part II consists of secondary roots. 



* Fdi/atu in Kellogg's Hindi Grammar, p. 202, § 377, is a misprint. 



t This influence of pdv even intrudes occasionally into the conjugation of jd 

 "go" ; thus the E. H. has sometimes ^ana "gone," like dvd, pdvd ; and the 3rd sing. 

 ■pioa.jdve is rather common in H. H. beside or Jdi/e. 



X This form is quoted by Dr. E. Mitra from the Pingala in the Vocabulary appended 

 to his edition of the Sankshipta-sara. I have not been able to verify it ; but the form, 

 is not intrinsically improbable. 



§ It is just possible to connect dv with the Skr. root api-i ; thus 3rd sing, apyeti 

 Pr. appc'i or dpei or dvc'i (cf. Jcddum "to do" for hattum). H. 0. 4, 400 seems to refer 

 it to Skr. d-pad (or better d-pat ?). The Bangali uses an altogether different root, dis 

 or as. Beames, in his Comparative Grammar (III, pp. 44, 45) rightly refers this root, 

 as well as the Sindhi ach, to the Skr. root d-gachh {pi d-gam). Disaspiration of an 

 aspirate and pronunciation of chh as s are not uncommon in the Indian vernaculars 

 (see my Comparative Grammar, ^§ 11, 145, exc. 2). The root dgachh would become 

 in Pr. dachh (sec Dclius, Radices Pracriticae, pp. 69, 70) or dgachh ; by contraction in 

 I'-^'.ngali', the former would become as (for dcUh), the latter dis (for dichh). The root 

 d'ls might, however, bo also referred to tho Sanskrit root d-vis, 



r 



