1859.] On the Introduction of Writing into India. 141 



The chief difficulties in the pronunciation of the Veda are the 

 changes of the final and initial letters.* The pupils are instructed 

 in these euphonic rules independently (the S'iksha), but whenever a 

 difficult case of sandhi occurs, the Guru examines his audience and 

 explains the difficulties. And here the method followed is this. 

 After the Guru has pronounced a group of words, consisting of 

 three or sometimes (in long compounds) of more words, the first 

 pupil repeats the first word, and when anything is to be explained, the 

 teacher stops him, and says, " Sir."f After it has been explained 

 by the pupil who is at the head of the class, the permission to con- 

 tinue is given with the words, " Well, Sir." After the words of the 

 teacher have thus been repeated by one, the next pupil has to 

 apply to him with the word, " Sir. "J When there is no difficulty 

 the rule seems to be that the Guru says two words at a time, which 

 are then repeated by the pupil. If it is a compound, one word 

 only is to be pronounced by the Guru, and to be repeated by the 

 pupil. After a section of three verses has thus been got through, 

 all the pupils have to rehearse it again and again. When they have 

 mastered it, they have to recite the whole without any break, witli an 

 even voice, observing all the rules of sandhi, marking slightly the divi- 

 sion in the middle of compounds, and pronouncing every syllable with 



of ud u shy a deva, ud u u shy a deva. This repetition would not take 

 place in udv eti, because u is changed into v. If sarvodatta could mean a word 

 being wholly udatta, then u would be excluded, and the rule would refer to a 

 only. But sarvodatta means recitation when the accent is disregarded, and all 

 syllables are pronounced with a high tone. The Commentary construes the rule 

 differently. I construe ^ SpT^ ^T^TTlt4 I rff^T^ WT^TTTSf^Tfi 



\» 



* These are chiefly the change of a final m into Anusvara before r and the 

 ushmans ; the common sandhi of the ushmans ; the suppression of a final n ; 

 its transition into r ; its transition into a sibilant j the absence of sandhi 

 where ri follows ; the sandhi of r, and the hiatus. 



t The text is fa^T^I rj &C. 



% Here again I differ from the Commentator, who takes parasya as an adjec- 

 tive referring to etad, i. e. guroh. At the end of a half-verse, this address, bho ! 

 is to be dropped ; at the end of an Adhyaya it is optional. 



