372 
REVIEWS. 
together to constitute a single perianth. M. De Candolle does not 
explain how it is that the third lobe, which is developed in Cary a, 
answers to lobe 4 and not to 3, being posterior and opposed to the 
bract. He states, however, that lobe 4 exists often in the male 
flower, especially in C. olivaeformis. In brief, the perianth of the 
male flower of Juglans consists of lobes 1 to 6, adnate to the sub¬ 
tending bracteole; that of Cary a consists of lobes 1 and 2 only, 
similarly adnate. In the female flower of Juglans all the perianth- 
lobes of the male are represented. In that of Carya , lobes 1, 2, and 
4, with the adnate bracteole alone form the 4-lobed perianth. 
In the genus Jlatycarya (.Fortuncea , Lindley) we have a yet more 
simple floral structure. It is thus interpreted: the male flower 
consists of stamens inserted upon a disk in the axil of the bract, and 
is destitute of a perianth. In the female flower there are two lateral 
appendages adnate with the ovary, which is inserted in the axil of a 
free bract. These appendages are regarded as representing two of 
the lateral stamens of the male, the perianth being wholly abortive. 
An instructive account is given of the morphological relations of 
the septa, which are intruded into the originally 1-celled ovary of 
Juglandese. With respect to the characters which distinguish Carya 
from Juglans , besides those already referred to in the flower, M. He 
Candolle points out that the fruits of Carya differ in the two valves 
being opposed to the stigmas, not alternate with them as in Juglans , 
in the sutures of these valves being but slightly marked, and the 
surface usually smooth or finely striate. We shall be curious to 
learn how far the relation of the valves to the stigma is a constant 
distinction. At present we can scarcely concur in the propriety of 
generically separating Carya from Juglans, especially that, as M. De 
Candolle says, the number of lobes of the perianth, employed as an 
important generic mark, is almost always reduced towards the ex¬ 
tremities of the male catkins in Juglans —the terminal flowers having 
but a 4 or 3-lobed perianth, as in Carya. It is worth while, in con¬ 
nection with this reduction, to refer to the observations and reason¬ 
ing of M. Alph. De Candolle upon characters variable in the same 
specimen, and their application for systematic purposes, in the 
Memoir upon the Cupuliferae before referred to. 
About 40 fossil tertiary species of Juglandess have been de¬ 
scribed. Although a very large proportion of these are probably 
either not Juglandeae at all, or not specifically distinct, yet we may 
safely assign an important role to the group at this early period. 
At the present time M. De Candolle reckons but 34 species. He 
has had the opportunity of examining but three fossil fruits, and these 
he inclines to refer to Juglans rather than to Carya ; an item of evi¬ 
dence, by the way, favouring an Asiatic rather than an * Atlantis’ 
migration, as explanatory of the community of type in the Tertiary 
vegetation of Europe and that now existing in East America. 
Six illustrative plates accompany the Memoir. 
