INTRODUCTORY. 7 



The term of development to which we have just made reference often lies within 

 such definite limits that, once ascertained, we may know about what to expect in 

 future trials with the same birds. In the pigeons I find that the length of the course 

 to be run, although varying widely, is, on the average, cut shorter and shorter as 

 the crosses range from close allies to more distantly related species. 



The causes of infertility in crosses — or, more correctly, the causes of fertility 

 in lower and lower degrees as the divergence between the crossed species increases — 

 have yet to be investigated. What connection, if any, do such phenomena have 

 with the formation of so-called "pure" germs in hybrids? If "purity" is a thing 

 of degrees, as it most certainly is, do the degrees rise or fall with fertility? Further- 

 more, if degrees are so numerous as to appear to flow together, can a few cases of 

 approximate "purity" be claimed as a law? Do higher degrees differ in kind from 

 lower, and so justify the assumptions of "discontinuity," "mutation," etc.? If 

 two wide-apart degrees are found with no intergrades, how easy to conclude that 

 they are both "sports" without genetic connection. 



In dealing with such questions, the first requisite is material suited to give 

 definite answers. In this respect the pigeon group is an ideal one, for it is composed 

 of so large a number of closely graded species that it must afford some positive 

 evidence of "flowing degrees," if such exist, and at the same time give us the 

 directions of rise and fall in a considerable number of well-defined differential 

 characters. For the experimental side of the study we require not only species that 

 will cross with continued fertility, but also species that will cross with varying 

 degrees of fertility. Both are essential and both are represented in this group in 

 exceptional abundance. 



As there are between 400 and 500 distinct species of wild pigeons, and as nearly 

 every species may be presumed to be fertile with at least one other species, the 

 number and variety of fertile crosses possible must, at the lowest estimate, be quite 

 large. But the number is probably many times larger, for a single species may be 

 fertile with all or most of its congeners and, in lower degrees, with members of other 

 genera and even of other families. 



The common ring-dove, for example, is fertile not only with some dozen 

 other species of its genus {Streptopelia, collared doves), but also in lower 

 degrees with the typical turtle-doves, as two successful tests — one with the 

 European turtle-dove {Turtur turtur) .and another with the Surate turtle-dove 

 {Spilopelia suratensis) of Burma and the Malay Peninsula — sufficiently attest.^' 

 In still lower degrees it is fertile with the mourning-dove (Zenaidura carolinensis) , 

 with the white-winged pigeon (Melopelia leucoptera), and with many races of 

 domestic pigeons. 



How much farther the fertility of this species may extend remains to be tested. 

 Some other ring-doves, as the wild species of China and Japan (Streptopelia douraca) 

 and the red ring-dove {St. humilis) of the same countries, give evidence of equally 

 strong fertility in crosses. The Oriental or Japanese turtle-dove {Turtur orientalis) 

 promises an even wider range of fertility, as I have obtained one fertile hybrid 

 between it and Colmnba livia doviestica. This is, I believe, the most remarkable 

 case of fertility hitherto recorded. It would certainly be difficult to match it in 

 any other group of the higher animals. The offspring of this hybrid exhibit to 



2 " Also later found to be fertile with the Japanese turtle-dove (T. orientalis). — Editor. 



