Professor Buhler on the Sanskrit Unguals. 127 



occurs also in pi-dhana, pi+dha, pi-shtap. Instances of the 

 simple loss of the first lingual occur in the second persons 

 plural Atmanepada of the so called fourth and fifth Aorists 

 and of the reduplicate perfect. 



The termination of the second person plural Atm. aorist 

 is 1 dhvam,' which, if attached to a verb ending in a diphthong 

 or a vowel other than a, a, must become c dhvam/ e. g. stu 

 makes a-sto-dhvam, nf, a-ne-dhvam. The reason of this 

 change is, that, as the fourth aorist is formed by combining 

 the root with the unaugmented forms of the imperfect of 

 the verb as, * to be,' the termination dhvam stands for 

 sdhvam. This form being preceded by i, i, u, u, ri, ri or a 

 diphthong must be changed first to shdhvam and next to 

 ddhvam according to the above rule. The latter form 

 ddhvam becomes dhvam by the rejection of the initial d as 

 in shodha. The rule regarding the change of ' dhvam,' dhve 

 is less strict in the two other cases mentioned, but based on 

 the same principles as in the fourth aorist. For the details 

 compare Benfey Vollst. Skt. Gram. § 830, Bern. 2, § 847. 

 Bern. 2, § 850, Bern. 2, where the explanation of these forms 

 is also given. It ought to be observed that this change of ' sh' 

 to ' t' stands in strict analogy with that of 's' to ' t' e. g, 

 vas -f- syami makes vat-syami, avas + sit, a vat-sit, ukha -f sras 

 ukhasrat, parna + dhvas, parnadhvat,* as + dhvam, addhvam. 

 In either case the production of t and t is caused by the 

 disappearance of the sibilant element of the sh and s under 

 peculiar conditions. The second mode in which ' sh' causes 

 the production of Unguals needs little comment. The rule 

 is simply this, that 1 n' preceded by ' sh' becomes n under 

 the same conditions as if preceded by r, ri or ri. 



Thirdly, hard dentals preceded immediately by ' sh' whe- 

 ther radical or a substitute for 9, ch, x, 9c, or j, are changed 

 to the corresponding Unguals. 



* This ^yord is mentioned by ^k*;^ 1 ™ I, 2, 72, 



