600 Journal of Agricultural Research voi.xx.No. 8 



We observed a ratio of 17 mule-foot to 25 cloven-foot in the F 2 genera- 

 tion, while on this latter hypothesis the calculated results would be 



10.5 to 31.5. The ^^ = iito =3 ' 43 ' The odds a S ainst Avia- 

 tions as wide or wider are about 45 to 1. But if we again add the 

 results of von Dabrowa-Szremowicz to ours, the observed ratio is 22 to 

 29, where 12.75 to 38.25 is the calculated ratio. In these combined 



, ^ deviation 9.25 „, , , . , . . 



results the = = 4.43. The odds against deviations as wide 



error 2.09 ^ ^° & 



or wider are now about 350 to 1 . In both cases the monohybrid explan- 

 ation is much more satisfactory. Furthermore, on a dihybrid hypoth- 

 esis we should sometimes obtain mule-footed when F 2 cloven-footed 

 segregates are mated together. To test this, such matings were made. 

 Two of the three cloven-foot F 2 daughters of 9 2 ( 9 2e and 9 2f in 

 Table II) were bred to a cloven-foot F 2 son of 9 3 (<? 31", Table II). 1 

 One F 2 9 gave 5 F 3 cloven-foot (4 c? <$ + 1 9 ) and the other F 2 9 

 gave 13 F 3 cloven-foot (5 d cf + 8 9 9 ). Therefore, a total of 18 F 3 

 cloven-foot was obtained from F 2 cloven-foot segregates bred inter se. 

 In the F 3 generation two cloven-foot 9 9(9 2f-b and 9 2f-c) were mated 

 to their cloven-foot brother, tf 2f-a, and gave 4 and 8 cloven-foot respect- 

 ively. We may conclude that mule-foot and cloven-foot represent a 

 single allelomorphic pair, in which the syndactylous form is dominant and 

 the normal form is recessive, and that extracted recessives breed true. 



As is common among mule-foot swine, the fused phalanges may sep- 

 arate along the line of fusion as the animal becomes older and heavier. 

 This splitting was infrequent in the front feet, but was occasionally seen 

 in the hind feet. There was never any difficulty in classifying the syn- 

 dactylous and normal at the time of birth or when the animals were young, 

 for syndactylism is a distinct discontinuous variation from normal. 

 There is, however, some variation in syndactylism itself. Fusion may vary 

 from complete, with no trace of separation on the hoof, to a less perfect 

 fusion with two deep parallel lines of demarcation. The former condition 

 is characteristic of the front feet, while the latter is the usual condition in 

 the hind feet. In an examination of 17 F 2 mule-foot segregates, 14 

 showed complete fusion in the front feet, but 1 6 showed the deep lines 

 of demarcation on the hind feet. The factor for syndactylism acts differ- 

 ently on the front and hind feet. (See PI. 70.) 



The relation of black to red in swine has never been quite clear. It is 

 well known that Poland China or Berkshire mated to Duroc- Jersey 

 usually produces a tortoise-shell type of red and black, but the amount 

 of each color varies markedly. Sandy, yellow, cream, or even white may 

 be substituted for red in such crosses, as shown by Severson (8). Wright 

 (10) advanced a suggestive hypothesis that such a tortoise-shell type of 

 sandy colored hog with black spots was selected in two directions to give 

 the characteristic color of the Berkshire or Poland China, on the one 



1 The relationship of all animals recorded in this paper may be obtained from Table II, the original data. 



