370 
REVIEWS. 
Woriel. Of these, thus extending westward as well as eastward from 
the G-alapagos, some reach the Cape Yercl Islands, West Africa, 
and the Cape of Good Hope, others fall off in Asia or its islands. 
But deducting 17 cultivated plants, or weeds of culture, from the 
above, the remainder, Dr. Andersson shows, are found to be species 
common to both hemispheres. This does not, however, invalidate 
the preponderant evidence in favour of the Central American and 
West Indian type. With regard to the relations of the species now 
endemic, Dr. Andersson has some interesting observations in con¬ 
nection with the inquiry as to their probable origin. From a colla¬ 
tion of representative and analogous forms, he arrives at the 
conclusion, confirming that arrived at by Mr. Darwin from zoological 
evidence, that it may be reasonably assumed that the Galapagos 
have derived their species from the American mainland, and that 
through the course of innumerable centuries a constant divergence 
from parent types has been in progress, resulting in the peculiar 
modifications which characterize the Flora of the present period. 
XXXVIII. —Memoire stir la Famille des Ju glandees. Par 
M. Casimir De Candolle. Ann. Sc. Hat. 4e. Ser. xviii. 
Three months ago we had occasion to notice the essay of M. Alphonse 
De Candolle apropos of his revision of the Cupuliferae for the Pro- 
dromus. The memoir now before us, by his son, M. Casimir, con¬ 
tains a resume of various organographic and miscellaneous observa¬ 
tions, accumulated during a like revision of the Juglandese for the 
same work. It contains some valuable matter, scattered through 
40 pages, concerning this small Hatural Order, which is one of 
peculiar interest, from the narrow range of its species in respect to 
structure—from its serving, like another very restricted group, the 
Hamamelidese, as a connecting link between Orders of simple and of 
complex floral and carpological character—from the wide geographi¬ 
cal distribution of the species, and from the important part which 
there can be no doubt they filled in the Flora of Tertiary Europe. 
Of the affinities of the group, very little is said. There is not 
much, indeed, to say. Dr. Findley’s view, that they are related to 
the oaks and chesnuts, M. De Candolle rejects, on the ground of the 
simple leaves, divided ovary, and usually pendulous or laterally 
attached ovules of these genera. This suffices to preclude their being 
actually classed among Corylacese, which, indeed, Dr. Findley does 
not attempt, although M. De Candolle almost leaves it to be inferred 
that he does so. Plis words are, “ Cependant M. Findley (Veget. 
4e Kingd.) ies reunit a V alliance des Corylacees, qui comprend les 
“genres Corylus, F., Fagus, F., Castanea , Gaert., Quercus, F.” Dr. 
Findley includes Juglandese with Corylacese in his Quernales —the 
“ Quernal Alliance,” and says of their representative, “ If the walnut 
“ had a many-celled fruit and a cupule, there would be no very good 
