122 THE NORTH AMERICAN SPECIES OF DROSOPHILA. 
mutant races will convince one that the viability of a mutant race is, 
on the average, inversely proportional to the degree of its difference 
from the parent race. In other words, the greater the change pro- 
duced by the mutation the more likely it is to interfere with the proper 
functioning of the organism. An organism is an extremely delicately 
adjusted mechanism, and any random change in it might be expected 
to decrease the efficiency of the whole. Furthermore, the greater 
the change the greater is the damage that is likely to result. It fol- 
lows, then, from actual observation as well as from theoretical con- 
siderations, that mutations which bring about slight changes are least 
likely to be harmful and therefore are most likely to become incor- 
porated into the race. This gives an explanation for the observed 
fact that specific differences are usually slight ones. The reason that 
the observed mutational differences are often of much greater degree 
is very simple; when such changes do occur they are easily discovered 
and are convenient to work with, so that they are artificially selected 
and perpetuated even if their viability or productivity is inferior to 
that of the parent race. 
That species commonly differ in more respects than do mutant 
races must mean that they differ in more inherited factors than do 
mutant races. Mutant races are usually known to differ significantly 
from the parent race in only one or a very few genes. That species 
differ from each other in many genes is, in the case of Drosophila or 
any form in which fertile hybrids are not known, only an inference.* 
Such a situation would, however, be pretty certain to arise as a result 
of long-continued isolation. In the case of the species of Drosophila 
that have been studied, interspecific sterility constitutes an effective 
mechanism for bringing about isolation. As to the origin of the inter- 
specific sterility itself, we can only speculate until we know more about 
the mechanism whereby such sterility is now brought about. 
Species, then, differ from each other in many genes. The differ- 
ences, though numerous, are such that each produces only a slight 
effect on the organism. ‘These differences are of the same kind as are 
the mutational differences, and may be supposed to have arisen by 
mutation. 
The picture of evolution that this analysis leads to is in effect not 
very different from that which Darwin drew. Species change gradu- 
ally, by the slow accumulation of numerous slight mutational differ- 
ences. The possibility of a sudden change of considerable degree is 
always present, but will not usually be realized, because such a change 
will generally give rise to an imperfectly adjusted organism. 

* This point has, however, been demonstrated in the case of certain species hybrids among 
plants. Compare Baur (1919, Zeits. ind. Abst. Vererb., 21, 48-52) and Lotsy (1912, Zeits. ind. 
Abst. Vererb., 8, 325-333) on Antirrhinum; Wichler (1913, Zeits. ind. Abst. Vererb., 10, 177— 
232) on Dianthus; Kristofferson (1914, Botan. Notis., pp. 25-31, abstract in Zeits. ind. Abst. 
Vererb., 14, 34) on Viola; van der Stok (1910, Teysmannia, 21, 47—59, abstract in Zeits. ind. 
Abst. Vererb., 4, 153) on corn-teosinte hybrids; East (1916, Genetics, 1, 311-333) on Nicotiana. 
