H. F. Wernhant. 
ISO 
their poral dehiscence, appendages, &c. The position of the ovary 
relatively to the other floral organs has thus been ignored. 
On the lines we have laid down in our opening chapters, 
however, it is urged that the occurrence of epigyny in so extensive 
a group may not be so lightly set aside. Epigyny, we have already 
suggested, is the reflection of a broad general principle—the 
Calycifloral tendency—correlated with the principles both of 
economy and of adaptation to insect visits. Now the Geraniales 
represent a line of evolutionary advance, delineated for the most 
part by the former principle, that of economy; there is no trace 
of epigyny in the Geraniales. In the search for the sign of an 
archichylamydeous ancestral stock for the Vaccinioideae it would 
seem that we should attempt to find a line characterized by the 
tendency to economy, coupled with the tendency to epigyny ; the 
working of the former would account for the similarity between 
the Vaccinioideae and the rest of the Ericales in the numbers of 
the stamens and carpels. Such a line is indicated in the progress 
from Ranales to Rosales, and within the latter heteorgenous 
cohort itself. Most of the Rosales have two whorls of stamens, 
each isomerous with the corolla ; in many, again, the carpels also 
are isomerous, although in this regard the archichlamydeous 
cohort shews further advance in economy than the sympe¬ 
talous group now under consideration; for the gynaecium is 
often oligomerous in Rosales, the numbers 2 and 1 being of 
of relatively common occurrence. This condition is analagous 
to that which we found in comparing the Ericales with the 
Geraniales; and the analogy is curiously enhanced when we con¬ 
sider the evolutionary advance reflected in the zygomorphy found 
in the large rosalian order Leguminosae. 
The similarity in dehiscence and detailed structure of the 
anthers must be weighed in the same balance with the relative 
position of the ovary: and when we consider the extensive occur¬ 
rence of epigyny throughout the Angiosperms, and the limited 
occurrence and obviously highly specialized nature of the ericalian 
anther character, the former seems to completely overbalance the 
latter in the investigation of this broad question of phylogeny. 
Epigyny, in fact, is one expression of broad general evolutionary 
progress, while this peculiarity in anther-structure cannot, prima 
facie , be so regarded; and this if only for the reason afforded by the 
sum total of the species in which the two phenomena respectively 
occur throughout the Angiosperms, apart from any biological 
