Osborn, it may be added, is in accord with this opinion of the minor importance of mutation. Bateson (1922) spoke for many of the geneticists when he expressed the same uncertainty of the application of the laboratory data to species in nature, and recently Anderson (1928) concludes a field and genetic study of two species of iris with the statement that there is little in his evidence to support the mutation theory of the origin of species.
Now, the problem of species obviously goes back to the recognition of the factors which may affect the potentialities of genes. I have nothing to contribute on this aspect of the subject. It is to an increased knowledge of the physicochemical nature of protoplasm and of the gene, and to such experimental work as that of Muller and others on mutations affected by the introduction of measurable amounts of energy, that we must look for the explanation of the first step in the process of evolution. But granting that mutations do on occasion occur, we may present a body of new data to show that these laboratory mutants are precisely the materials which have differentiated many of the species of our genus Cynips.
This evidence becomes available because there are, in the family Cynipidae, more than 70 species of gall makers which have rudimentary or reduced wings strikingly different from the long wings normal among the other seven or eight hundred described species of the family. The differences between these two types of insects are illustrated on several of the plates accompanying this paper. Material on all of these 70 species will be brought together in a later study, but it may be said that all of the data support and extend that part which is here presented on the genus Cynips.
The typical wasp of the family Cynipidae has wings which are somewhat longer than its body. The wings even approach twice the body length (a wing-body ratio approaching 2.00) in certain genera of Cynipidae. The normal wings vary between the different genera and subgenera, but are remarkably constant among the individuals of each taxonomic group. Thus the normal ratio is always about 1.35 in the subgenus Atrusca, 1.50 in the subgenera Besbicus and Cynips, and 1.60 in the subgenus Antron. On the other hand, many Cynipidae have wings less than one-fifth the body length (ratios under
