CERTAIN CROSSINGS OF JAPANESE DOVES. 25 



only 5 hatched; 14 showed no development; 2 formed embryos. One of the young- 

 was alive at 2 years or age, and another escaped when 8 months old. The other 

 3 lived less than a month. 



The sex of 16 offspring of this series of matings, involving T. orienialis in a cross 

 oi family rank, is known. Of these 14 were known to be males; possibly 2 were 

 females, one having a very rudimentary ovary, the other copulated as a female 

 (a very inadequate test of sex) and almost certainly produced no eggs, though it 

 lived nearly 5 years. 



The data on the fertility of the reciprocal cross arc not vciy extensive, but 

 indicate complete infertility. Three such pairs of d" orienialis x 9 common pigeon 

 gave 43 tests, all of which proved absolutely infertile. These wholly infertile tests 

 are further analyzed as follows : 



Orienialis (19) X gyrans (1903) Fam 10. ... (F 22) 



Orientalis (1) X gyrans (1) Fam 12. ... (C 7/0) 



Orientalis X domestioa (black) Fam. . . . (15?). . . (F 22) 



Fertility Tests of Hybrids op Common Pigeon X Japanese Turtle-dove. 



Three males from pairs 1, 5, and 8 described above were tested for fertility. 

 Only one of these (A 1) — the one that probably lived longest and arose from the 

 most fertile original cross (pair 5) — proved fertile. This male was mated back to 

 a common pigeon; * the restrictions upon the fertility and upon the life-term of 

 offspring of this cross (pair 9) are notable and are presented fully in table 6. 



An examination of the record makes it clear that though some development 

 proceeds from some of the fertilizations of this pair, it is, in a very high percentage 

 of cases, quite narrowly limited. There is a high percentage of eggs (28 of 59) 

 showing no development whatever; 13 produced embryos only; 17 hatched, 1 or 2 

 of these requiring to be helped from the shell; 6 of those hatched died within 2 to 

 27 days. The length of life of 5 young is not known, but probably only 4 lived 

 to maturity. This very bad record may have been somewhat influenced by the 

 element of inbreeding, since the hybrid was mated to a probable half-sister; but 

 that this is by no means the chief adverse factor is indicated by the fact that the 

 short period during which an unrelated female {C. affinis) was used there was still 

 less fertility — no trace of development in any of 5 eggs. A factor of more impor- 

 tance, perhaps, is that of overwork. This hybrid and his consorts were worked more 

 rapidly, during much of this mating period, than is favorable for the production of 

 strong germs. Nevertheless the main features of this mating record is clearly to be 

 referred to the hybrid nature of the sire, as will become clear later when the fertility 

 of hybrids from closely related species shall have been examined. The failure of 

 germs to begin development, the limitations placed upon the development and upon 

 the life-term of the offspring, all become more and more i^ronounced as the hybrids 

 are derived from more distantly related species. 



From this cross more of the second eggs of the clutch showed the less development 

 of the pair — 11 to 6. One female of this cross was of white color (color of her 

 paternal and (?) maternal grandfather) ; two females had the color of the father, 

 and one resembled the mother. One male ^\as white, one was chequered, and one 

 was black. 



* A female C. livia was also used during a short period; this jiair proved entirely infertile. 



