260 APPENDIX III LETTERS PUBLISHED IN NATURE. 



class, it is due to diversity in the action of life-preserving habits while 

 at the beginning of the process competing with each other; if under 

 the third class, it is due to diversity in life-preserving habits while not 

 competing with each other. 



Now, in some of the cases in the second class and in all those of the 

 third class, it is impossible that the differences should be useful. This 

 is most easily shown as regards the third class ; for if in any case a new 

 character attained by one of the sections is an advantage, then the 

 same character would be an advantage for each of the other sections, 

 exposed to the same conditions in other regions, and, therefore, there 

 is no advantage in the difference. 



If my thought is correct, some of the differences produced by diver- 

 sity in the action of the several forms of reflexive segregation and selec- 

 tion, and all those produced by diversity in the action of environal 

 selection, when that diversity is due to different habits that are not 

 necessitated by any difference in the environment, are non-useful dif- 

 ferences. Therefore, besides the principle of " correlated variation " 

 referred to by Professor Lankester (Nature, vol. LIV, pp. 245, 365), we 

 have other explanations of certain kinds of specific characters that are 

 not useful ; but the class of characters of which right-handedness and 

 left-handedness are examples seem to lie beyond the reach of these 

 explanations, and perhaps beyond the reach of the explanation sug- 

 gested by Professor Lankester. 



5. Letter by T. D. A. Cockerell, with Suggestions on the Facts Mentioned Above. 

 The following letter in reply to the above appeared in Nature for 

 May 13, 1897: 



THE UTILITY OF SPECIFIC CHARACTERS. 



Under the above heading, in your issue of April i, Mr. J. T. Gulick has an inter- 

 esting communication, in which he asks whether it is possible to explain right- 

 handedness, the dextral or sinistral coil of snail shells, and similar features, as 

 having any utility of which they are certainly characteristic. Can it be due to 

 natural selection that one snail is dextral while another is sinistral? 



It is a curious fact, I think, first pointed out by Mr. Call, that in the American 

 fresh-water shells of the genus Campeloma, sinistral shells are more numerous 

 among the young than among the adults. Thus, for example, Mr. H. A. Pilsbry 

 (Nautilus, February, 1897, p. 118), states that Miss Jennie F. Letson examined 

 a lot of Campeloma desisum for him, with the result that out of 68 1 specimens, 

 mainly adult, but including those from one-fourth grown up, none were sinistral. 

 Out of 410 shells of the uterine young 3 were sinistral, slightly over 0.73 per cent." 

 He adds: "Probably all who have collected Campelomas have noticed the greater 

 proportion of sinistral examples among the young shells. This doubtless indi- 

 cates that the reversed condition is an unfavorable one for maturition." 



So here, at any rate, we have some direct evidence as to selection. I think it 

 will strike anyone that while left-handedness might be as good for the race as 



