19171 
BRYAN—HABENARIA 39 
Vol. II, Dudley writes that in many specimens of Н. hyper- 
borea flowers have been found without spurs and with the 
labellum like that of H. dilatata. 
The question arises: Are our new varieties produets of 
spontaneous variation (mutability), or have they been evolved 
from the ever-present fluctuating variations which a species 
offers as material for natural selection to work on? Accord- 
ing to the latter and more conservative hypothesis the extreme 
variation, namely, var. ecalcarata, fig. C, owes its existence 
to the **gradual summation of small deviations in one direc- 
tion, through succeeding generations," the intermediate 
formal variety varians, fig. B, indicating a transitional form 
in the series. These small deviations (fluctuating variations) 
being useful, according to the theory, offer the essential ma- 
terial which natural selection has gradually accumulated in 
one direction, resulting in an extreme type like our variety 
ecalcarata. This slow formation of species would necessarily 
require a long period of time. The existence of intermediate 
forms, like our formal variety varians, would seem to furnish 
proof for this conservative belief in the slow formation of 
species. 
On the other hand, variety ecalcarata may be regarded as 
a probable mutant from H. psycodes; and the formal variety 
varians may then be a hybrid between the species and the 
extreme variety. An experimental proof could be attempted 
in this ease. In favor of this view it may be said that the 
sudden origin of new forms other than by a series of tran- 
sitional stages is in accordance with the facts of plant breed- 
ing. De Vries derives the doctrine that variability may be 
increased by selection; one of the chief objects of his book, 
he says, is **to try to show that ordinary or fluctuating vari- 
ability does not provide material for the origin of new 
species. He speaks of Ше ‘‘illusion of an increase in vari- 
ability." The existence of intermediate forms, according to 
the eonservative view, is usually pointed out as filing the 
gaps between the discontinuous series that species form. De 
Vries says, however, that these are not transitional forms, 
but are independent types, which he calls elementary species 
