ee GENERAL REMARKS. 
Manchuria, Shantung, Kiangsu, Chekiang, Fokien, Formosa, Pesca- 
dores, Kuangtung, Hongkong, Hainan, Yunnan, Szechuan, Kansuh, 
Hupeh, Kiangsi; and as regards other countries, for Corea, Japan, 
Luchu Islands, Siberia, India, etc. This is most valuable informa- 
tion. One of the greatest merits of the Index, however, is the 
reduction of a large number of synonymic, or else useless, deter- 
minations. For example, Rhododendron indicum, Sw., is now relieved 
of 25 botanical names. Descriptions are confined to new species 
and to a few new genera. As no key is given to the Chinese 
genera of each order and to the species mentioned in the Index, 
and no complete botanical library is available in China, this work 
cannot assist to determine any plant unknown to us, though it 
may be mentioned in the book. The aim of the Index is another. We 
need along with it a short Handbook of Chinese Botany, or BENTHAM’S 
Flora Honghongensis enlarged to A Flora Sinensis. Introductory | 
matter could be reduced and description be made as short and 
characteristic as possible. iN 3 
Classification will remain partly unsatisfactory till all plants of the 7 
globe are known and time can be given to closer observation and more 
minute investigation of every species in a genus and of every genus — 
in an order. Soil, elevation, moisture, light and air in their innumer- 
able differences and varying combinations are so many conditions” 
which modify the typical form of a plant. Cultivation has shown 
the great extent of variation in some species. Observation of wild 
Species is limited. Many differences, now regarded as sufficient to 
establish a distinct species, may mark only variations. Genera, t00, 
with only one or two species attached to their fame and fate, will ba 
recognized by a critical specialist as extreme forms of some - 
genera. Very few of the Chinese species are in every respect identical 
with specimens of the same species in Europe or America. The present 
difficulties of determining plants in China would be considerably 
removed if the above-mentioned Desideratum could be realised. 
