30 INHERITANCE , FERTILITY, AND S^EX IX PIGEOXS. 



The parents of pair 3 were certainly not closely related, probably not related 

 at all. They were both long-lived, but the female of this pair died a few months 

 after the close of the season recorded in table ISa. They were not overworked. 

 6 of their 7 eggs hatched; 1 showed no development. The long-lived offspring were 

 from May and June. The first egg and the last egg hatched the shortest-lived young. 



Pairs 4 and 5 were most adequately tested and are the matings of greatest 

 interest. The same female was used in the two matings; the members of pair 4 

 were probably not related; in pair 5 the female of the former pair was mated with 

 her son. This record covers a period of 8 years. A juvenal male offspring {o6) 

 from pair 4, from the third year of this mating, is shown in pi. 2. The most im- 

 portant feature of these two matings is the appearance of three "mutations" at 

 the close of the period, when the dam was at least 12 to 14 years old and when 

 inbred to her son. The distribution of sex, longevity, and fertilitj', as displayed 

 in table 19, is also of much interest. 



In a closer examination of these two matings we note : 



(1) Two sexes arose from the same clutch in 14 cases; in 10 of these the first 

 egg produced a male, in 4 cases the first hatched a female. 



(2) In 9 of these same cases the bird from the first egg lived longer than its 

 clutch-mate from the second egg; in 4 cases less long; in 1 this is unknown. 



(3) Of the 9 long-lived birds. 7 are males, 2 are females; of the 4 short-lived 

 birds from the first egg 2 are females. 



(4) Two females hatched from mdividual clutches in 4 cases; in two of these 

 the bird from the first of the clutch lived much the longer. 



(5) Of 7 such pairs of males the birds from the first egg Hved longer in 5 cases. 

 Birds from the first egg therefore hved longer than their mates in 16 such cases, 

 died earlier in 8. at the same time in 2; the data are insufficient in 5 cases. 



{&) A tabulated summary of the longevity of offspring, fertility, and the sex- 

 ratio corresponding to each of the 7 years this female (of pairs 4 and 5) was mated, 

 is as follows: 



Average age (14) yg. of 1903 = 42+ mo.;" eggs unhatched=0 (in 19): Scf ; 5 9 



" (12) " "1904=23+ •• " " =1 (in 16) ; Tcf ; 6 9 



" (9) " "1905=18+ " " " =2? (in 14); 4(^:4 9 



" (S) " "1906=24+ =1 (inl2); 5o';49 



" (S) " "1907=18+ =? (inl4); 4<f;49 



Inactive 190S; mated with son, 1909-10. 



Average age (2) yg. of 1909= Si mo.; eggs unhatched = 1 (in 4); lcf:291 .. 



>o mutations 

 " (G) " " 1910=26+ " " " =2 (in S): 2d'; 39/ 



The longest term of life (1903) coincides with the highest proportion of males 

 and with the highest fertihty; the very shortest term of fife (1909) coincides with 

 the maximum of infertility, the highest proportion of females, with very old age 

 of the mother, and to inbreeding with an immature son. . 



(7) White color " mutations '' occur only at the end of the series, when the female 

 parent is quite old (12 to 14 years in captivity), when fertihty is lowest, when the pro- 

 portion of females is highest, when longevity is reduced, and when inbred to her son. 



The question of the nature and breeding value of the color "mutations" which 

 appear in this record, and of the effect of inbreeding of Japanese turtles, on the 

 fertihty, sex, longevity, and the general topic on "strength of germs," will be 

 considered in the next chapter. 



' In these series which include one or more birds alive at the time of writing, a plus sign (+) is added to the average 

 aee (in months). 



