REVERSION IN GUINEA-PIGS AND ITS EXPLANATION. 5 



agouti young can scarcely be questioned, since he was himself an agouti. The 

 total progeny of the 24 males by red females is 133 red, 66 black, and 69 agouti, 

 the expectation being 134 : 67 : 07. By agouti mates they produced 25 red, 

 21 black, and 63 agouti young, the expectation being 27 : 21 : 61. These are 

 remarkably close agreements. One of the 24 males, 7932, was also mated 

 with a black female, producing 2 black and 2 agouti young, the expected 

 equalit} r ; no red young were to be expected. 



Agouti animals produced by mating agouti animals inter se are not so 

 uniform in behavior as those thus far discussed. Experiment shows that they 

 fall into four groups: 



I. Agoutis producing red, black, and agouti young. 

 II. Agoutis producing only red and agouti young. 



III. Agoutis producing only black young and agouti young. 



IV. Agoutis producing only agouti young. 



Group I is represented by nine agouti males, whose parents were both 

 agoutis; they were enumerated in the third division of table 2. Together they 

 have produced (a) by red mates, 34 red, 20 black, and 19 agouti young; expected 

 36 : 18 : 18; (b) by agouti mates, 3 red, 5 black, and 14 agouti young; expected 

 5 : 4 : 12, if the females were of the same character as the males, as most of 

 them doubtless were; (c) by black mates, 3 black and 3 agouti young, the ex- 

 pected equality, no reds being expected. 



Group II is represented by the ten animals enumerated in table 3. They 

 proved to be incapable of producing black young in any sort of mating. Mani- 

 festly they were homogyzous in agouti. By red mates they produced 53 red 

 and 52 agouti young. They were accordingly heterozygous in red, and equality 

 of red and agouti young was to be expected. By agouti mates they produced 

 13 red and 62 agouti young. The expectation in this case varies with the 

 character of the agouti mate employed, which was not in every case definitely 

 determined. If the agouti mate belongs in Group I or II, then the expec- 

 tation is 1 red : 3 agouti; but if the mate belongs in Group III or IV, the 

 expectation is all agouti young. If the mates are a mixture of the two sorts, 

 as a random group of F 2 agoutis should be, then the expectation lies between 

 75 and 100 per cent agoutis; the percentage obtained is 82.66. By black 

 mates, the Group II agoutis produced 18 young, all agouti, as expected. These 

 were sired by two animals, both of which by red mates had produced red 

 young as well as agouti young. 



Group III is represented by the 16 tested animals enumerated in table 4. 

 They were incapable of producing red young, and hence were homozygous in 

 black. But since they produced black young as well as agouti, it is evident 

 that they were not homozygous in agouti. Three females (374,437, and 462) 

 are included only provisionally in this group, since they were not tested by 

 matings with red animals, but only with black ones. They are known to have 

 produced black young, but it is not certain that they could not also have pro- 

 duced red young, in which case they would fall in Group I. The test in the 

 case of several other animals (as 5, 68, and 314) rests upon too small numbers 



