24 GENETIC STUDIES OF RABBITS AND RATS. 



length of the hind-leg in rabbits, but it is evident from table 32 that 

 the action of this factor is general, not local, and that it affects the 

 general framework of the rabbit, not its leg-length alone or its hind-leg 

 rather than its front-leg. 



In the size of the soft parts of the body — as, for example, ear- 

 length and body-weight — females in general surpass males. But it is 

 possible that this difference would not hold for certain small races of 

 rabbits. We have already noted that in our pure Polish rabbits the 

 average weight of the males was actually greater than that of females, 

 although in all the large races and in all our cross-breds, females are 

 heavier than males, the difference amounting to about 5 per cent. 

 In ear-length there is very little if any difference between the sexes. 

 Males have slightly longer ears among the Polish-Himalayan cross- 

 breds, but females among the Flemish cross-breds. The diflFerences 

 are of doubtful statistical significance. 



The differences in size due to sex are too small to affect materially 

 the correlation coefficients obtained by comparing bone-measure- 

 ments, ear-length, and weight, as in tables 12 to 30. Of this I con- 

 vinced myself by making separate correlation tables for the two 

 sexes in the case of weight correlated with length of tibia, in which 

 the disturbance due to sex would be at a maximum, since males and 

 females would in these two characters diverge most and in opposite 

 directions, males having a longer tibia, but females being heavier. 

 The correlation between weight and tibia based on both sexes is 

 0,758d=0.015 (table 20). For the males alone it is 0.745=b0.023; 

 for the females alone it is 0.775±0.021. The correlation is not in- 

 creased by tabulating the sexes separately. For the sexes combined 

 the correlation is practically the mean of the values obtained for the 

 sexes separately. Neither sex departs more than the probable error 

 from the combined value. 



TESTS FOR LINKAGE WITH COLOR OR OTHER COAT 



CHARACTERS. 



Hoshino (1915) demonstrated the existence in garden peas of a 

 strong linkage in heredity between a quantitatively varying char- 

 acter, time of flowering, and red color of the flowers. It would be of 

 interest to know whether in animals a similar linkage exists between 

 large or small size and any particular color gene. I can not discover 

 that such is the case in rabbits, probably for the reason that size in 

 rabbits is determined by genes located in many or all chromosomes, 

 whereas each color gene is located in a single chromosome. In each 

 of the two crosses with Flemish, a different albino allelomorph was 

 introduced by the small-sized parent. Albinism was completely 

 recessive in Fi, but reappeared in F2 in approximately one-fourth of 

 the individuals. If any linkage existed between albinism and small 



