108 



THE AMERICAN NATURALIST [Vol. LI 



in following: a cross with the wild male, was young of mean grade 

 3.05. Seven of these of mean grade 2.95 produced a second F, contain- 

 ing hooded young of mean grade 3.47. This is a reversed regression of 

 0.52 on the grade of their actual hooded grandparents, or of 0.42 on 

 the group from which their grandparents were chosen. Their mean lies 

 about midway^ between that which would have been expected from 

 the original hooded female (5513) had no crossing with wild rats oc- 

 curred and that which was observed in the first F^. 



Obviously these facts do not harmonize with the assumption that 

 the regression observed in the first F, was due to loss of modifying fac- 

 tors accumulated during the ten preceding generations of selection; for 

 no further loss occurs in the second F,. On the other hand, a par- 

 tial recovery is made of what was lost in the first F,. This suggests 

 the idea that that loss may have been due to physiological causes non- 

 genetic in character, such as produce increased size in racial crosses; for 

 among guinea-pigs (as among certain plants) it has been found that F^ 

 has an increased size due to vigor produced by crossing and not due to 

 heredity at all. This increased size persists partially in F^, but for the 

 most part is not in evidence beyond F^. I would not suggest that the 

 present case is parallel with this, but it seems quite possible that similar 

 non-genetic agencies are concerned in the striking regression of the first 

 F, and the subsequent reversed regression in the second F^. 



gression in a second F, is very clear, as other cases than those already 

 discussed will show. 



A hooded rat of grade +4 and generation 10, J* 6348, had by a wild 

 female several young of the character already described for the young 

 of $5513. These, mated brother with sister, produced a first F, (table 

 141) of 90 rats, 22 of which were hooded, 68 being non-hooded, again 

 a good 1 : 3 ratio. The hooded young ranged from + 2 to -}- 4 in grade, 

 their mean being 3.28. Of the 22 hooded individuals, 1 male and 7 

 females were mated with wild rats to obtain a second F., and the 

 second F^ animals were then mated brother with sister to obtain the 

 desired second F,. The character of this is shown family by family in 

 table 143. It contained 497 individuals, of which 121 were hooded 

 and 376 non-hooded, a ratio of 1 : 3.1. The weighted mean of the 8 

 selected grandparents is 2.93, which is 0.35 below the mean of the 22 

 first F, hooded animals which they represent. The mean of the second 

 F, hooded young is 3.22, which indicates a reversed regression of 0.29 



sin The Scientific Monthly (Jan. 1916) I have stated that a second 

 cross showed "a return to about what the selected race would have been 

 had no crossing at all occurred." This is obviously inaccurate and should 

 be corrected. It rests on a comparison with the combined average of both 

 the older and the more recent experiments. [MacDowell devotes half a 

 page to demolishing the statement already corrected here.] 



