30 



The Journal of Heredity 



A photographic chart showing these 

 crosses given in Figure 12. 



The number of chicks expected from 

 this cross on the above assumption 

 compared vi^ith the number obtained is 

 given below : 



Black Columbian Buflf 



$ &9 5 & ? 2 



Obtained 40 33 6 



Exj)ected 39 30 10 



The agreement of the actual results 

 with those expected is good, if one ex- 

 cepts a slight deficiency of buffs and a 

 slight excess of Columbians. With the 

 small numbers involved these depart- 

 ures are not serious. 



When the first generation Duckwing 

 male is crossed with buff (double re- 

 cessive) females equal numbers of all 

 classes of chicks should appear. The 

 results compared with this expectation 

 follow : 



Obtained. 

 Expected. 



Black Columbian Buff 



$ &9 S &9 9 



19 12 



20 



10 



10 

 10 



The first generation black females 

 (from Columbian 9 x black S ) were 

 crossed with Columbian and with buff 

 males. On our hypothesis they should 

 be heterozygous in extension and 

 should not transmit the gene for sil- 

 ver. The results are given in Table 2, 

 crosses 7 and 8. By a Columbian 

 male they produced equal numbers of 

 black and Columbian chicks of both 

 sexes; (15:14) by a buff male they 

 produced equal numbers of black and 

 buff chicks of both sexes (16:13) : 



Finally, the first generation Birchen 

 females were tested by crossing with 

 Columbian and with buff males (Table 

 2, crosses 2 and 3). By hypothesis 

 they should transmit extension to half 

 their progeny and silvering to their 

 sons only. All daughters should be 

 buff or black carrying buff. The cross 

 with the buff male is here the critical 

 test. 



From this cross resulted the follow- 

 ing: 



Black Black 



(silv'd) (non-silv'd) Columbian Buff 



$ $ sex? sex? 9 sex? 



Obtained 3 2 2 2 4 2 



Expected 3 4 4^ 4 



No exceptions to expectation were 

 found and the agreement is as good as 

 could be expected with such small 

 numbers. 



Considering all of the crosses to- 

 gether, it is found that all support the 

 hypothesis proposed, as far as down 

 color of chicks is concerned. The cross 

 of extended black X restricted gave in 

 the first generation thirty-two extend- 

 ed. The backcross of the first hybrid 

 generation with restricted (buff or 

 Columbian) gave ninety-nine extended 

 and ninety-eight restricted — equality 

 expected. Extended black appeared to 

 be completely dominant to restriction 

 and completely epistatic to silver in 

 the down pattern. The segregation of 

 silver and buff must then be measured 

 in the restricted classes of chicks and 

 here the crosses of all heterozygous 

 Columbians bred in terse or crossed 

 with buff produced a total of forty- 

 seven Columbian and twenty-two buff, 

 as compared with forty-four Colum- 

 bian and twenty-five buff expected. 

 Silver in all cases segregated as a sex- 

 linked character dominant over buff. 



Due to a high mortality from an 

 epidemic disease, many of the back- 

 cross chicks recorded did not live to 

 develop adult plumage. Those which 

 did survive are recorded under "adult 

 colors" in Table 2. All of the black 

 chicks as adults developed plumage in 

 which black was extended to other 

 parts of the plumage in addition to 

 wings, tails, and hackles, while the buff' 

 and Columbian chicks developed adult 

 plumage in which black was restricted 

 to wing, tail, and hackle. The amount 

 of black in these parts was variable. 

 In some chicks from the backcross 

 with Columbian, the Columbian pat- 

 tern segregated in its original form 

 (Fig. 13) both on silver and buff' 

 ground colors ; while some were ob- 



