LANMAN. — PALI BOOK-TITLES. 671 



titles (Dlgha, etc.), but also very commonly by the specific title of the 

 sutta in question: thus, at i. 982, Brahmajale (=D. I. 1) ; at xxi. 654, 

 Potthapada-suttante (=D. I. 178). — Of the Khuddaka-nikaya, he cites 

 Udana and Niddesa and Patisambhida-magga by name (the last, twenty- 

 five times), but none of the rest except Sutta-nipata and Jataka. These 

 he cites not by the general title, but by the special title of the sutta or 

 Jataka-tale in question : thus, in quoting a couplet from Sutta-nipata 

 i. 8 3 , he says (ix. 36) simply yarj ca Metta-sutte . . . tiadi vuttarj. It is 

 most instructive to note that in citing, for instance, the Ratana-sutta, 

 he cites it (at xiii. 166) as a well-known and much-used text (a paritta), 

 and does not care whether we think of it as constituting Sutta-nipata 

 ii. 1 or as Khuddaka-patha vi (ed. Childers, p. 6 or 314). — Of the 

 Abhidhamma books, he cites four, Vibhanga, Katha-vatthu, Dhatu- 

 katha, and Patthana. The Vibhanga he cites perhaps 14 oftenest of all, 

 and mostly by that title, but sometimes by chapter-titles, thus, iii, iv, 

 and v, as Dhatu-vibhanga, Sacca-vibhafiga, Indriya-vibhanga. — Typ- 

 ical forms of Buddhaghosa's citations are the following: "Such indeed 

 is the opinion of the Dlgha- and Sarjyutta-professors. But the Majjhima- 

 professors will have it that ..." Idarj tava Dighabhanaka-Sarjyutta- 

 bhanakanarj matarj: Majjhimabhanaka pana . . . icchanti (viii. 952). 

 Similarly, Evarj tava Dighabhanaka: Majjhimabhanaka pan' ahu . . . 

 (viii. 1179). Evarj tava Majjhimabhanaka: Sarjyuttabhanaka pana 

 . . . ti vadanti (xiii. 541). 



Requirements for a good system of citations. — The essential parts 

 of a citation are two, — the title of the book and the indication of the 

 place in the book. The requirements of a good system are ready intel- 

 ligibility, brevity, convenience, and precision. The first three concern 

 especially the abbreviations of the titles, and the last two concern the 

 indication of the place. Moreover, to be readily intelligible, the ab- 

 breviations must be unambiguous and easily remembered. It is evi- 

 dent that the citations of the ancients fail to meet most or all of these 

 requirements. And, as appears in the sequel, the like is true of the 

 abbreviations that have hitherto been in use among Pali scholars. I 

 have good reason to hope that the designations here proposed will 

 prove to be so suggestive and so easily remembered as to win general 

 acceptance. 



Indication of the place in the book. — This is a subject which I 

 should like to discuss at length if it were not so hopeless. An extreme 



14 This Vibhanga is liable to confusion with the Vibhanga of the Vinaya 

 and is in fact so confused in an Index of Proper Names in the Visuddhi-magga 

 made, I presume, by an amanuensis of H. C. Warren. 



