Campanulatce. 299 
So long as Sympetalae and Archichlamydeae are separated 
sharply as distinct groups, Cucurbitaceae are entitled a place among 
the former. But this place, we conceive, must be relatively 
isolated, in view of the considerations to which we have referred 
above; and this family cannot be regarded even as emerging from 
the line which led to the Campanal Stock, but must rather be con¬ 
sidered as the immediate sympetalous derivatives of those groups of 
Engler’s Parietales which exhibit the tendency to epigyny. 
Summary of Infers. 
It may be well at this point to summarize briefly our study of 
the Infei 'ie, illustrating our summary with the aid of the evolutionary 
tree figured on p. 230 in the preceding chapter. 
We have recognized, in the short sketch of the Archichlamydeae 
furnished in Chapter II, a definite tendency to epigyny along certain 
lines of descent, and we have suggested that certain biological 
advantages are attendant upon this character. In the endeavour 
to secure these advantages, if we may so express it, the several 
floral forms and groups have doubtless made not a few experiments 
in the course of descent; in other words, epigyny may be the out¬ 
come of more than one morphological modification. One, the 
progressive hollowing of the thalamus and ultimate fusion of the 
latter with the carpels, is clearly recognizable in a series of forms. 
There are, without doubt, others; and epigyny, like angiospermy 
and zygomorphy, has very probably been attained in more than 
one way. But the data provided hitherto by research into floral 
development are, unhappily, so meagre, that any detailed discussion 
of the matter, invaluable in more fortunate circumstances, must 
inevitably be barren and misleading in the present state of our 
knowledge. 
We must be content for our purposes, then, with the observed 
fact of epigyny, however produced, and with a knowledge of the 
extent of its occurrence among groups which are to be regarded as 
natural apart from considerations of the relative position of the 
ovary in the flower. 
Beside its very probable origin in more ways than one, we find 
the occurrence of epigyny in more than one undoubted circle of 
affinity. In some groups— e.g., Ranales, Sapindale,—epigyny 
is so rare that it may represent merely the special tendency 
expressed in a solitary genus ( Eupomatia) or strictly within the 
