195 
But the book is most noteworthy for other reasons. It is the 
first which a student of the flora of any province of China can 
use to identify its plants, for it is provided throughout with keys 
to families, genera, and species. Indeed, no similar work has 
yet been completed for any other part of the Far East. А quite 
unique feature is the combination, in certain cases, of genera of 
similar habit in a single key.* 
The largest family is Leguminosae with 66 genera and 173 
species; then follow in order Gramineae, 76 genera, 166 species; 
Cyperaceae, 21 genera, 126 species; Compositae, 50 genera, 117 
species; Orchidaceae, 44 genera, 89 species; Euphorbiaceae, 33 
genera, 88 species; Rubiaceae, 34 genera, 86 species; Urticaceae, 
25 genera, 70 specics; these families taken in the wide sense. 
The nearest extra-Chinese regions to Kwangtung are Formosa 
on the east, the Philippines on the southeast, and Indo-China on 
the southwest. With the last the transition in floras is probably 
very gradual, as continues to be the case through Tonkin and 
Annam into Cochin-China and Cambodia. There is also great 
similarity between the floras of Kwangtung and Formosa, many 
of the differences being due to the high mountains of the latter. 
'The distance from eastern Kwangtung to the nearest point in 
the Philippines is only two thirds of the width of the Chinese 
province, or one third the length of the Islands, but the flora of 
the two political divisions is quite different. For example, the 
numericial order of phanerogamic families in the Philippines is 
Orchidaceae, Rubiaceae, Leguminosae, together having about 750 
more known species than Kwangtung; pteridophytes, also, are 
three times as numerous. An area less than twice as great can- 
not accountfor this; the explanation must be sought in higher 
mountains, and tropical profusion. 
It has been found necessary to describe only 15 species as 
new. This is partly due to earlier publications by the authors 
themselves as well as by others, for Hongkong alone has 100 
endemic species; in part it must be attributed to the fact that 
[* By a incidence, and in a widely separated locality, the same pro- 
cedure has just PHONE followed in the recent admirable flora of the pine barrens 
by Witmer Stone.—Eb.] 
