ICFFECT OF SEASON OF HATCHING UPON DISTRIBUTION OF FERTILITY. 61 



of "crowded reproduction." The St. alba male (715), which was tested with the 

 above female, was produced under similar seasonal conditions as the inadequate 

 male 421; but the former was not an inbred. When bred to his sister a con- 

 siderable proportion of the eggs developed '» (table 32). The trace of fertility 

 found in 421 is, however, the only trace that will be found among the 7 known 

 males of the series. 



Male 464 proved wholly infertile in 6 successive tests with a St. alba, as is 

 shown in detail in table 33. Similarly male 440 was as completely infertile with 

 St. alba in 13 successive tests (table 34). The complete infertility (in 5 tests) of 

 the male last to be considered (446) is recorded in table 35. In this case the female 

 {St. alba) used in the cross was tested immediately before and immediately after 

 her infertile union with 446. The earlier mating was with a T. orientalis male, 

 a brother to 446, but from the previous year; in the later mating a male St. alba 

 was used. In both of these matings the tests proved fully fertile. Male 446 was there- 

 fore wholly responsible for the 5 infertile eggs produced during his mating period. 



It will be of interest to present here the record for 1913 of the female used in 

 the three matings just described in connection with the fertility tests of male 446. 

 In the first place, this will better show how completely fertile was the female with 

 which that bird was so completely infertile; and secondly, the data will be seen to 

 reflect, in the offspring of overworked St. alba parents, some of the features which 

 until now we have considered in T. orientalis only. 



The data recorded in table 36 bring out the following points: 



(1) The fertility of 31 eggs was tested. The first 22 of these were fully fertile; the 

 twenty-third and twenty-fourth were able to reach full embryonic de\'elopment but unable 

 to hatch; the "first" egg of the following clutch was fully fertile, while its mate produced 

 only a four-day embryo (imperfect shell may or may not have caused this). The following 

 clutch contained an infertile and a broken egg; the next an infertile and a fully fertile egg 

 which produced a bird, probably weak; it died 10 days after hatching, and presented at 

 that time no well-defined sex-glands. The last of the 31 eggs tested for fertility produced 

 a 6- to 8-day embryo. A total of 51 eggs were produced during the season. Those not 

 accounted for in this table were used in other work (chemical analj^sis). 



(2) Seven of 16 eggs laid before July 1 and incubated, produced males; 3 produced 

 females. After July 1, 3 males and 4 females were hatched. Though these numbers are 

 not large, they are in both cases sunilar to those of the previous year (table 35). 



(3) The two clutches which immediately precede the first unhatchable clutch both 

 produced males from the first and females from tlie second egg. 



(4) The sex-glands of a 1-month-old bird from the end of the season (September 23) 

 were largely or wholly undeveloped. 



(5) Though the data for length of life are not yet complete, many birds being still 

 alive, it is clear that the later eggs of the season gave rise to birds of shorter life-terms. 

 The "seasons" in the previous year (table 35) also bear a similar relation to longevity in 

 addition to sex, as noted above. 



(6) These results obtain from a mating that approximates to a mating of two indi- 

 viduals of the same species, the female being a f alha x \ risoria hybrid. It is evident 

 that the great decrease in fertility at the end of 1913 was largely due to reproductive 

 "overwork," though the predominance of males at the begimiing of the season and the 

 general vitality of the young were possibly somewhat infiuenced by the slight element 

 of crossing which is also involved. 



'"This second male also died of tuberculosis on March 1, 1915. 



