PROF. T. H. HUXLEY ON THE GENTIANS. 115 
genious speculation. On the contrary, everything confirms it. 
As I have already remarked, the principle which underlies the 
progressive modification of the flower, in each of my series, is 
the gradual interposition of more and more obstacles in the way 
of easy access to the nectaries. Yet these obstacles, whether 
they arise out of a mere narrowing and elongation of the tube 
of the corolla, as in Lissanthe, or out of the development of 
fimbriz and pectines (as in Lophanthe and Stephananthe), or of 
interlobes, synanthery, and discoidal stigmata (as in Ptychanthe), 
are always of such a nature that they afford no effectual hindrance 
to the passage of a thin and flexible organ, like the haustrum of 
a suctorial insect, but only just sufficient obstacle to make the 
insect an efficient agent in cross-fertilization. 
Few inquiries would be more interesting and profitable 
(though undoubtedly few more difficult) than the working out 
of the geographical distribution of the Gentianes in relation 
to the Insect Faun of the several regions in which they occur. 
I merely note that those Gentianex which have remarkably long 
infundibulate corolle are found in regions, such as Madagascar 
and Guiana, which are tenanted by large Lepidoptera provided 
with long haustra. 
Müller's doctrine respecting the origin of the forms of the 
flower in the Alpine Gentiane appears to me to be equally appli- 
cable to the whole of the Order, and to supply a true cause 
whereby the morphological facts may be correlated and explained. 
I do not suggest that the interaction of cross-fertilizing 
insects with the variation in the structure of the corolla 
accounts for all the characters by which the five hundred spe- 
cies of Gentianex differ from one another. Why one species is 
annual and another has a rhizome; why the great majority have 
opposite palmati-veined leaves, while some have alternate or 
pennati-veined leaves; why the majority have a single-celled 
ovary, and yet, in many, the ovary is more or less completely 
two-celled* ; why most are herbaceous, while some are bushy aud 
* "The partially, or wholly, two-celled ovary appears to be generally regarded 
as the result of the gradual introrsion of the margins of the carpellary phyl- 
lomes into the cavity of a primitively one-celled ovary. But had “Haplanthe” — 
a one-celled ovary ? Must not the apocarpous condition precede the syncar- 
pous? Does not the state of the ovary in Apocynee and Asclepiadex rather 
suggest that the primitive Gentians (or perhaps, I should rather say, the primi- 
tive stock of all these orders) had an ovary like that of a hypogynous Saxifrage ? 
