388 Statistics of the Flora of the Northern States. 
pe: P 
but less isolated species, the supposition of a double birth-place 
for the exceptional species would be the most natural; although 
one would im be inclined to regard them as mostly cases of 
closely related species whose points of difference are still unascer- 
tained or undervalued, For we no more know how nearly alike 
two species may appear and yet be specifically distinct, than we 
know how widely they may differ and yet own a common origin. 
tanist’s best conclusions regarding the limitation of spe- 
cies are seldom more than judgments on imperfect data, con- 
stantly liable to be questioned and revised. 
But both these most striking cases, and the transitional ones 
between them and those of ordinary distribution, are becom- 
ing too numerous to bear this exceptional mode of explanation. 
DeCandolle lays much stress upon the isolated occurrence of 4 
single peculiarly United States species, Phryma leptostachya, 0 
Nepal. Now the foregoing catalogue includes three or four addi-, 
tional cases of the same kind, which Drs. Hooker and Thomson s 
Himalayan collections may probably double; and the consider- 
able number of North American species which meet Himalayan 
ones in Japan already indicates the line of connection between 
these two distant floras. We should therefore look in one and 
the same direction for the explanation of these enn | 
no less than of the more ordinary cases of distribution, 
adopting the conclusion to which DeCandolle himself arrives, 
and maintains on various and convincing grounds,—Viz., 
plants must have been created at different epochs, and that the 
greater part of the existing species are older than the present 
configuration of our continents,—should refer such anomalous 
distribution to very ancient dispersion; and all the more confi- 
dently as the known examples of the kind increase in number. 
