4 REVERSION IN GUINEA-PIGS AND ITS EXPLANATION. 



reds which I have had, derived from other sources than these two, have failed 

 to produce agouti young in crosses with black animals; see the summary in 

 table 1, p. 8. They may therefore be considered to lack altogether the agouti 

 factor in which wild species of cavy are regularly homozygous. 



The agouti factor is transmitted through albinos exactly as are other color 

 factors. Accordingly an albino may or may not possess and transmit the 

 agouti factor; if it does possess this factor, it will produce agouti offspring 

 when crossed with a black animal. Several examples of this have been ob- 

 served among the albino descendants of cf 2054. Thus red c?3496, table 1, 

 which when mated with heterozygous black females produced red young and 

 black ones but no agoutis, when later mated with the albino females 4270 and 

 4262 produced 4 red and 4 agouti young. The father of these albinos, it should 

 be said, was an agouti animal. 



Agouti animals produced by crossing black with red do not breed true. 

 They produce three classes of young if bred with each other or with ordinary 

 reds. The latter form of test has been more commonly employed in my experi- 

 ments because I had more red animals than agoutis and it was easier to ascer- 

 tain their gametic composition. Fourteen young, however, were obtained from 

 the mating of Fi agoutis inter se; they were 2 red, 2 black, and 10 agouti; see 

 table 2, p. 9. The mating with reds produced 111 red, 45 black, and 40 agouti ; 

 expected, 98 : 49 : 49. This expectation is calculated as follows: It is expected 

 that half the young will be red because the agouti parent is known to be hetero- 

 zygous for red. It is also expected that half the young will receive the agouti 

 character, which those possessing black pigment will show, but which those 

 possessing only red pigment will not show. Hence all the reds should look 

 alike, but half the others should be agouti. In the mating inter se of agoutis 

 of this generation the expectation is 4 red : 3 black : 9 agouti ; the observed 

 numbers, as stated, are 2 : 2 : 10, in a slightly smaller total, 14. The tests 

 sununarized above relate to five different jPi reversionary agouti males. Three 

 of them were descendants of my original red male, 2054; the other two (9150 

 and 9152) were sons of Horton's red female. All gave similar results. 



Agouti young produced by the mating already described of an Fi rever- 

 sionary agouti with an ordinary red produced the same kinds of young as did 

 the Fi agoutis, viz, red, black, and agouti; see the second division of table 2. 

 The tests applied in this case were identical in character with those applied to 

 the previous generation, viz, matings inter se and with ordinary reds (lacking 

 A). Twenty-four different male animals were tested in one or both of these 

 ways, while several females were tested in the first-mentioned way. The data 

 for these are recorded for convenience in connection with the male involved in 

 the same experiment. Only two of the various animals tested failed to pro- 

 duce all three classes of young. In the case of these two, 9602 and 9685, there 

 can be little doubt that a more extended test would have produced the missing 

 class. Male 9602 produced 2 black young and 1 agouti. A further test would 

 almost certainly have given red young, since his mother was red. Male 9685 

 produced 4 red young and 2 black ones. That he could also have produced 



