66 THE ORCHID REVIEW. [Marcu, 1914 
genotype must be non-heritable modifications.’”’ But we should say there 
is much virture in that IF, and we do not think that species and varieties 
-are going to be got rid of it in this free and easy fashion. 
It may be difficult to give an exact and comprehensive definition of a 
‘species, but the difficulty will not be got over by calling it a homozygote. 
And we cannot wait while experiments are undertaken to show whether the 
hundreds of thousands of recognised species faithfully reproduce their kind 
for ever. And there are organisms that are homozygous in one character 
and heterozygous in another. How are these to be defined under the new 
scheme? But we do not admit that species are mere abstractions. Are 
bium nobile and Cypripedium insigne abstractions because they vary ? 
Dendrobium aureum is found on the lower slopes of the Himalayas, in 
‘Ceylon, in the Philippines, and in the island of Celebes, and is not 
absolutely identical in each separate area, but it is none the less a well- 
‘defined species in the ordinary acceptation of the term. And it would be 
absurd to say that any differences that it presents inthese different areas are 
due to hybridisation. Without attempting an exact definition, and without 
requiring that they shall reproduce faithfully their kind for ever, we venture 
to assert that species exist, and that they can be recognised by their 
-characters and behaviour. 
Species originate, increase and multiply, and diffuse themselves 
from their original birthplace, so far as climatal and other barriers 
permit. And in gradually extending their area they come under the 
influence of changed conditions, and, being plastic, they respond to the 
environment and, variation resulting, ultimately give rise to what are 
known as geographical races, or representative species, by the survival of 
the variations best fitted for the new conditions. Every species occupies 4 
definite geographical area, over which it extends more or less continuously 
according to circumstances, and it is a fact well known to naturalists that 
the most closely allied Species mainly occupy distinct geographical areas. 
The changes may be, and undoubtedly are, very gradual, but such evidence 
establishes the case for the existence of inheritable variability. Crossing 1S 
undoubtedly a potent cause of variation, but to say that it is the sole cause 
is only to ignore the evidence of geographical distribution. 
Hybrids are far more common in nature than many of the older 
naturalists were willing to admit, and the fact has been chiefly demonstrated 
by numberless horticultural experiments, which, however, were largely 
undertaken with other objects in view. The behaviour of oe also, is 
understood. sufficiently to show that it is not parallel with the phenomena 
