400 



SCIENCE 



[N.S. Vol. XXIX. No. 740 



in a heterozygous (XT) germ cell. In this case 

 T would become the " synaptic mate " of T, and 

 X would be left once more (as originally) 

 without synaptic mate, fit instrument for the 

 origin of new progressive variations, the char- 

 acters determined by Y now being the com- 

 mon property of both sexes. 



A clue to phylogenetic histories would thus 

 be afforded us, giving point to such variations 

 as the lugens variety of Ahraxas. Thus it is 

 possible (though nothing but pure speculation 

 in the light of our present knowledge) that 

 lugens maybe the phylogenetically older form, 

 characteristic originally of both sexes, and 

 that the grossalariata character may have had 

 its beginning in gametes lacking the X-ele- 

 ment, i. e., in a T-element formed as the 

 " synaptic mate " of X. Thus would arise 

 grossulariata males, but the new character 

 being dominant over its antecedent (lugens) 

 would quickly be transferred to the females, 

 since these contain a no-X (i. e., a T) ele- 

 ment, in common with the males. But the 

 X-element, as shown by Doncaster's experi- 

 ments, is still unassociated in a gamete with 

 the new grossulariata character, and so the 

 fixing of that character upon the species is not 

 yet complete. 



How now may the occasional reappearance 

 of lugens females be accounted for? Simply 

 by reduction divisions, in spermatogenesis, in 

 which the two Y-elements fail to segregate as 

 normally, forming in consequence a sperm- 

 cell which lacks T (the grossulariata char- 

 acter). If such a sperm-cell fertilizes an egg 

 of the constitution LX, a lugens female is 

 certain to result. 



If, as has been suggested, the presence in 

 one gamete and absence from another pro- 

 duced by the same cell-division, of an " odd 

 chromosome " (or other X-element, whether 

 chromosome or something else) is itself a 

 circumstance which favors the origin of new 

 characters in the defective (male-forming) 

 gamete, then we shall perhaps come to attach 

 less importance than has sometimes been done 

 to the supposed influence of sexual selection 

 in evolution. For sexual selection, as has 

 often been pointed out, can in no case account 

 for the origin of new characters, and it is 



extremely doubtful whether it plays any part 

 even in their preservation. 



Striking new characters produced by in- 

 ternal causes doubtless persist unless sup- 

 pressed by external causes, i. e., unless they 

 disqualify their possessor for competition in 

 the struggle for existence. There is no more 

 reason for supposing that males gain their 

 gay colors and markings from choice on the 

 part of the females, than that females owe 

 their modest colors to choice on the part of 

 the males. But if, as suggested, the very 

 mechanism of gametogenesis is adapted for 

 the production of new characters in the male, 

 then we are afforded a basis for their expla- 

 nation, without invoking external causes. 

 Recent investigations tend strongly to show 

 that variations of evolutionary significance 

 are primarily internal. This is unmistakably 

 so in the matter of sex. Even in cases where 

 sex is subject to control by environmental fac- 

 tors, as in aphids and daphnids, the environ- 

 ment acts indirectly apparently through the 

 control of the same internal factors which 

 govern sex in other animals. If the mechan- 

 ism which I have suggested is not their true 

 source, then we may well look for other pos- 

 sible internal mechanisms. 



Orthogenesis also, the persistent tendency 

 of an organism to vary in a particular direc- 

 tion, irrespective of the action of natural 

 selection (if indeed orthogenesis be a reality, 

 which, however, I do not assert), orthogenesis 

 then would find an explanation along simUar 

 lines to those which I have suggested. For 

 if a Y-element arose because of the very lack 

 of X, then it would be natural for it to con- 

 tinue to grow until it became the full comple- 

 ment of X. 



I make no apology for offering the hypoth- 

 esis, or hypotheses, contained in this paper. 

 I would have every reader recognize as fully 

 as I do that they are hypotheses, and I shall 

 be quite content if they suggest lines of in- 

 vestigation which will further elucidate the 

 nature of sex and the manner of its inherit- 

 ance. W. E. Castle 



Zoological Labokatobt, 



HaBVAED UNrVEESITT, 



February 10, 1909 



