284 Segregation and Species [ch. 



portfolios of a herbarium, or drawers of butterflies and 

 moths to discover abundant *' species" which are analytical 

 varieties of others. The principles of heredity we trace in 

 our experimental breeding are operating throughout the j 

 natural world of species. They may apply to the inter- j 

 relations of allied forms which are species in the strictest 

 acceptance of that term. Remarkably clear and un- 

 controvertible evidence of this fact has been obtained by 

 Dr Ezra Brainerd in his studies of the American Violets. 

 The whole series of observations is full of significant details 

 which I must be content to omit from this summary. The 

 essential facts are as follows. Finding wild in nature plants 

 which from his knowledge of the genus Viola he determined 

 as accidental hybrids between certain species, he removed 

 these plants to a garden and kept them under observation. 

 They proved to be very nearly, but not quite completely, 

 sterile, thus manifesting one of the attributes of crosses 

 between genuine species in the strictest sense. Some small 

 quantity of seed was however produced in the cleistogamic 

 capsules, and therefore may be taken as certainly the result 

 of self-fertilisation. This seed gave rise to plants showing 

 obvious segregation in regard to many features — the colour 

 of the stems, the fruits, and the seeds, and also, though not 

 quite so palpably, in the shapes of the leaves. In some at 

 least of these derivative plants there was — as Herbert and 

 others have found in similar studies of the offspring of sterile 

 hybrids — some return of fertility. Such an experiment raises 

 the hope that successful investigation of the nature even of 

 the sterility consequent on crossing, the most obscure of all 

 genetic phenomena, may become one of the possibilities of 

 Mendelian research. 



It is scarcely necessary to insist that plenty of the 

 characters which are now known to segregate would be far 

 more than sufficient to constitute specific differences in the 

 eyes of most systematists, were the plants or animals in 

 question brought home by collectors. We may even be 

 certain that numbers of excellent species universally recog- 

 nized by entomologists or ornithologists, for example, would 

 if subjected to breeding tests be immediately proved to be 

 analytical varieties, diftering from each other merely in the 

 presence or absence of definite factors. 



