Mutations and Evolution. 
75 
few, we have already seen that the behaviour of the others can 
probably be accounted for on the basis of sterility or lethal factors, 
explanations which now enter into many features of the CEnotheras. 
Again, the duplication of a chromosome in CE. lata and other forms 
is paralleled by the phenomena of non-disjunction in Drosophila , 
although there are certain differences which will be pointed out. 
But in Drosophila the rapidity in breeding large numbers in many 
generations has made it possible to carry the analysis of the germ 
plasm much further than in any plant. 
In the single species Drosophila melanogaster, the “vinegar 
fly,” over 150 mutations have been observed. These include more 
than 25 factors for eye colour, and many more for body colour and 
form of wings, while many structural differences are also involved 
in other mutations. The majority of these have appeared hut 
once. Some mutations are, however, more frequent. White eyes, 
a mutant from the wild red-eyed fly, are known to have occurred 
three times in the New York experiments (Morgan, 1919, p. 248) 
and several times in the cultures of other observers. Vermilion 
eye-colour has appeared at least six times, “rudimentary” wings 
five times, cut wing four times, truncate wing frequently but 
probably owing to different changes. As with the (Enothera 
experiments then, certain mutations repeat themselves, hut 
apparently with much less frequence. This suggests that in 
(Enothera the mutation often appears much later than the germinal 
change (pre-mutation) which gave rise to it, being suppressed in 
many cases by lethal factors causing sterility. 
A number of cases of probable reversion from the mutant 
character to the original wild type have occurred, but most of these 
are uncertain on account of the possibility of contamination. In 
the experiments of May (1917a) with bar-eye, however, this 
possibility is excluded. This character is dominant to normal 
(round) eye, and the reversal takes place with sufficient frequency 
to eliminate the possibility of error. Reverted individuals give 
only normal offspring. Bar-eye differs from the normal in having 
fewer facets, and the stock showed considerable variability. In an 
experiment in selecting for more or fewer facets, May obtained 
six full-eyed males and five females heterozygous for full-eye. They 
were indistinguishable from normal and were explained as simple 
reverse mutations. In a later paper (May 1917b), however, a more 
complex explanation of this reversal is favoured. It is assumed 
that the normal wild fly carries a limiting factor with respect to 
