method of transcription adopted by standard works in English dealing
with each, for French and German transcriptions, whatever their merits
may be as representations of the original sounds, are often misleading
to English readers, especially in Chinese. For Chinese I have adopted
Wade's system as used in Giles's _Dictionary_, for Tibetan the system of
Sarat Chandra Das, for Pali that of the Pali Text Society and for
Sanskrit that of Monier-Williams's _Sanskrit Dictionary,_ except that I
write ś instead of s. Indian languages however offer many difficulties:
it is often hard to decide whether Sanskrit or vernacular forms are more
suitable and in dealing with Buddhist subjects whether Sanskrit or Pali
words should be used. I have found it convenient to vary the form of
proper names according as my remarks are based on Sanskrit or on Pali
literature, but this obliges me to write the same word differently in
different places, e.g. sometimes Ajâtaśatru and sometimes Ajâtasattu,
just as in a book dealing with Greek and Latin mythology one might
employ both Herakles and Hercules. Also many Indian names such as
Ramayana, Krishna, nirvana have become Europeanized or at least are
familiar to all Europeans interested in Indian literature. It seems
pedantic to write them with their full and accurate complement of
accents and dots and my general practice is to give such words in their
accurate spelling (Râmâyana, etc.) when they are first mentioned and
also in the notes but usually to print them in their simpler and
unaccented forms. I fear however that my practice in this matter is not
entirely consistent since different parts of the book were written at
different times.

My best thanks are due to Mr R.F. Johnston (author of _Chinese
Buddhism_), to Professor W.J. Hinton of the University of Hong Kong and
to Mr H.I. Harding of H.M. Legation at Peking for reading the proofs and
correcting many errors: to Sir E. Denison Ross and Professor L. Finot