514 



THE AMEBIC Ar^ NATURALIST [Vol. LV 



prising regularity. Thus for instance, in the strain 

 London, the earlier generations (120 female mice) showed 

 a tumor incidence of 27 per cent., in the intermediate 

 generation (61 females) the figure was 38 per cent, and in 

 the later generation (197 female mice) 28 per cent. Sim- 

 ilar conditions were found in a number of our strains. In 

 certain strains, however, variations* in the tumor rate 

 did occur. While some variations may of course be ex- 

 pected, in case the number of mice considered is very 

 small, there occurreji in addition changes which can not 

 be attributed to this factor. 



In the majority of cases in which these latter changes 

 did occur in our stock, they consisted in a decrease in 

 the tumor rate in later generations ; in a few cases only 

 there occurred an increase in the tumor rate. These 

 changes were in all probability due to two factors: (a) 

 In certain families and strains as a result of long con- 

 tinued inbreeding a gradual decrease in fertility and 

 vigor occurred. Associated with this change was in cer- 

 tain cases a noticeable decrease in the tumor rate. Es- 

 pecially in the strain No. 8 there seemed to be a connec- 

 tion between loss in resistance to disease and fertility and 

 the decrease in the tumor rate. This strain was inbred 

 for seventeen generations and the changes in the tumor 

 rate seemed to occur step by step in correspondence with 

 the progress in inbreeding. Under those conditions the 

 connection between inbreeding and change in the tumor 

 rate appears the most probable explanation, although it 

 can not be considered as definitely proven as yet. 



(b) Various factors caused a selection to take place 

 within the strain ; certain families died out, while others, 

 which happened to be more resistant to a certain disease, 

 survived, propagated and thus gained a preponderance. 

 These surviving families differed sometimes in appear- 

 ance, oi- li! viu'oi-. ill the 1)rli;i\ior towards certain 

 lable tuiiH.rs. Such clinimv- ncr'c accompanied 

 cases a cliaimv in the liiiiior rate. In the majo: 

 our cases a decrease occurred ; in a few cases a 



