168 MUTATION 



tion of mutually adapted sex organs may be 

 readily conceived as follows: Let a new species 

 differ from its ancestor by a character m; then, 

 in accordance with the familiar fact of sex dimor- 

 phism in xmit characters this takes in the two 

 sexes the forms m' and m". If the two sex-forms 

 are incompatible the new species will come to 

 nought; but if not incompatible the two modi- 

 fied sexes may interbreed and be prevented from 

 breeding with the parent species. By the addition 

 of a series of new unit characters, n, o, p, etc. — 

 each of which must stand the test of compatibihty 

 in the two sexes — a complex dimorphic organ may 

 be built up by mutation. It were wearisome to 

 attempt to catalogue the mutant-like variations 

 that have been recorded among insects and other 

 animals. The great work of Bateson, Materials 

 for the Study of Variation^ is full of instances, 

 and the entomological and conchological jour- 

 nals are full of many more. Everywhere we find, 

 along with the universal quantitative variation, 

 cases of quaUtative, discontinuous variation in- 

 volving entire unit characters; and these new 

 characters are, probably, judging from our expe- 

 rience with domesticated animals, inheritable. 



MUTATIONS UNDER DOMESTICATION 



When we study a group of domesticated or- 

 ganisms, such as poultry, we find the races dis- 

 tinguished by characters that do not intergrade 



