270 THE TREND OF THE RACE 
we have made practically no progress in ascertaining, in any form, 
whatever relation may exist between the nature of the variation 
and the kind of external stimulus by which it is evoked. 
A large part of the congenital variations that appear in organ- 
isms are mere products of the mingling of factor differences con- 
tained in the germ cells of the parents. Where such variations 
are not obviously the expression of typical Mendelian inheritance 
they are frequently explicable as unusual factor combinations 
which are nevertheless essentially Mendelian. Certain variations 
may perhaps be attributable to the loss of factors and others to 
the reduplication of one or more factors, as the result of some 
anomalous behavior of the germ plasm, such as occurred in the 
mutant Gnothera gigas and several other similar cases. But all 
such variations as these are probably of minor significance in 
relation to the general problem of progressive evolution. They 
are the results of the shuffing of the cards, and at best they can 
produce only new combinations of old factors. 
There are writers (Lotsy, Hagedoorn) who hold that the kinds 
of variations just alluded to are the only ones of which we have 
any evidence. But if we admit the existence of this kind of 
variability only, we are landed in serious difficulties. There is 
certainly no adequate reason for denying that variation is a real 
phenomenon dependent upon qualitative changes in the germ 
plasm. Many cases are known in which the appearance of new 
mutants is in all probability dependent upon such qualitative 
germinal variations. But with few exceptions their occurrence 
seems entirely fortuitous and we can form no conjecture as to 
their possible cause. 
There is a certain amount of experimental evidence that 
germinal modifications may be evoked by environmental agen- 
cies. The experiments of Tower on the production of mutants in 
the Colorado potato beetle, and the work of MacDougal and 
Gager in inducing mutations in @nothera and other plants by 
salt solutions and radium are among the few investigations on 
multicellular organisms which have yielded positive indications of 
germinal response to changes in the environment. 
