236 Scientific Intelligence. 
vées, by DeCandolle, a notice of which is still due to our readers. 
Among the interesting matters contained in the Introduction, we 
note the statement of the author that cultivation, even where 
immemorial, has in no wise effaced the limit of species. =A. G. 
3. The Colors of Flowers ; by Grany Atten. Macmillan & 
Co. 1882.—This taking little book, in which a popular subject is 
very interestingly treated, originated as an article in the Cornhill 
Magazine, was extended into a series of articles in Mature, and 
now the latter are reédited and collected in a volume of Nature 
Ser We need only notice the distinguishing feature of a con- 
tribution to evolutionary science which must already have been 
widely read. The “central idea” is that petals are transformed 
stamens, rather than transformed leaves. ‘The argument is, that 
the earliest flowers consisted only of stamens and pistils, one oF 
oth; that the original color of these was yellow, that conse 
quently (by inheritance) the stamens of almost all flowers are 
ellow,—whence “it would seem naturally to follow that the 
earliest petals would be yellow too.” Now “the earliest and 
. 
simplest types of existing flowers [i. e. the petals of such as have 
suppose no convincing proof is to be had from observation. Yel- 
so more commonly may be the anther, but rarely the filament, the 
dilatation of which is assumed to give rise to petals:—to give 
rise, moreover, to the sepals also, if the theory holds, at least 
when they are colored. ut how when they are green an her- 
baceous? And how is the line to be drawn; and if colo 
sepals originated from stamens, why not subtending bracts 8 
well, when these are petaloid ? 
Our author says: “We can see how petals might easily have 
taken their origin from stamens, while it is difficult to undel- 
stand how they could have taken their origin from ordinary 
leaves, a process of which, if it ever took place, no hint now 
mains to us.” But either we have a hint in the brilliant bracteal 
