LANMAN. — PALI BOOK-TITLES. 687 



typographically awkward. 31 Thus the V. V. of the Lists will at once 

 be condemned by any one who has the " typographic sense ; " and sim- 

 ilarly A. A. (cf. List 10) for Anguttara-atthakatha. 



The capitalization of the second of the initials representing the two 

 members of a compound is unnecessary, and, as increasing sensibly the 

 number of obtrusive characters on the printed page, gives to it an 

 unrestful effect (macht das Satzbild unruhig : Baensch-Drugulin). 

 This effect is aggravated by the interposition of a period between the 

 two. Examples : P. V., K. V., P. P. ; so P. V. A., K. V. A., List 4. 



Comments on the abbreviations in their order. — Much of Avhat is 

 to be said in justification of this or that abbreviation has already been 

 said by way of illustration of this or that canon. What remains I give 

 in the order of the abbreviations concerned (as they appear in Table 

 I) and with references to previous discussion. 



Comment 1. The uniliterala. — D. and M. and S. and A. for the 

 first four Nikayas. Other things being equal, the fewer the letters, the 

 less suggestive is the abbreviation. Hence the class of uniliterals 

 should be kept within the very narrowest limits. They are in fact so 

 few and have to be used so often, that they will be easily remembered. 

 To maintain their efficiency, abbreviations, like domestic plumbing, 

 should be used constantly. 



Comment 2. The biliterals. — Texts of the Khuddaka-nikaya. 

 Although the diaskeuasts have grouped the Stanzas of the Male Elders 

 and the Stanzas of the Female Elders separately as Thera-gatha and 

 Therl-gatha, the two texts are so truly one 32 that they should cer- 

 tainly be designated by the same letters, Th. The difference is most 

 clearly and unobtrusively indicated by an appended Arabic 1 and 2 ; 

 and is so indicated in fact by Davids and Carpenter in the edition of 

 Sumangala-vilasini, p. xvii. The like applies to the texts of the Expo- 

 sition, Major and Minor ; it is far more practical to have the differen- 

 tiated term, Niddesa, come first, and the differentials last 33 (especially 

 since those differentials are Maha and Culla: see p. 681). Even Nd. m - 

 and Nd. c - or Nd. ma J- and Nd. ,nin - are better than MN. and CN. ; but 

 Nd.l and Nd.2 are better still. 



31 Or can be read as an ill-sounding or unpleasantly suggestive combina- 

 tion. To me, at least, the abbreviations Thag. and Thig. (sic) have always 

 suggested cruel Thugs and Dacoits rather than gentle Theras and Therls. 



" 32 Observe that Buddhaghosa (D.cm., I. 15) says Vimana-peta-vatthu 



Thera-therl-gatha, treating these four texts of the second pentad as two groups. 



33 This principle is duly recognized by the administrations of the great 



metropolitan post-offices. Thus we have "London EC," "Berlin SW.," not 



"EC. London," "SW. Berlin." 



