330 



THE AMERICAN NATURALIST [Vol. XL 



lected bv Dr. W. A. Xason); the shells are much larger than 

 those of Fi<;. 1, and measure 30 mm. in length. It will also be 

 noted that the \ ariation in this lot of shells is not so marked as in 

 those illustrated in Fig. 1. There is considerable difference 

 between the extremes but the majority of specimens conform 

 more or less to a single type, the large, corpulent form. A study 

 of several hundred specimens from each locality shows that each 

 lot varies a certain percentage toward a given form. In lot 1, 

 (Fig. 1), the variation is 75 percent toward no. 1, while in lot 2 

 (Fig. 2) the variation is about 60 percent toward no. 8. Without 

 more data to disprove it, this would seem to point to the fact that 

 the species of each locality varies toward a definite form. In lot 

 no. 1 the dominant form is palnstris, while in lot no. 2 the domi- 

 nant form is nufialliana. 



The interesting fact in connection with all this is (and this is 

 where de Yries's mutation theory seems applicable) that all the 

 forms illustrated will develop from the same egg capsule. The 

 eggs laid by nuttalliana will produce narrow palustris as well as 

 the fat parent form, while the narrowest palustris will likewise 

 produce the fattest nuttalliana. May this not be an illustration, 

 also, of two types (although this subject is treated under ever- 

 sporting varieties by de Vries) which he calls poor races and rich 

 races. In Fig. 1, 75 percent of the progeny are palustris (assum- 

 ing that the parent was a palustris form) and are of the rich race, 



while in Fig. 2 (as- 



^sm. suming the parent to 



^9fk be palustris) the prog- 



I eny are of the poor 



3 Vxniti nin V ilv it i 1^ V trUvrinata Sa.y fomi. Only 40 percent 

 2 r trirariiialn rn„fus„ Walkrr. V . himrinata being the pttlustris 



Lea. 4. r. Walk.r. All enlarged. This, hoWCVCr, 



