78 The Mechanism of Evolution in Lbptinotaksa 



the cross of L. signaticollis and L. diversa, giving a simple monohybrid reaction, 

 with no interchange of characters, while both species in other combinations 

 show interchange of characteristics, and L. diversa and L. decemlmeata form 

 at the other extreme a fixed combination, composed of elements from both 

 parents which it has not been possible to dissociate. 



It therefore appears to me that capacity for dissociation is not necessarily a 

 property of the basal species complex, but rather due to the reactions to which 

 the parental gametic masses are subjected, both in the way of the combination 

 in which they are united, and also depending upon the conditions of the medium 

 under which the imion or reaction is carried out. In this there is no difference 

 in principle from the union of non-living materials where the nature of the 

 reaction is entirely a product of the combination made, and the conditions under 

 which the reaction is set to operate. 



This capacity of the two gametic systems to unite for the production of a 

 heterozygote, acting in harmony for the production of a soma, with two parental 

 groups of potentialities which attain a balance for the space of a single genera- 

 tion, but with no permanent capacity for union, with the total separation of the 

 parental systems in gametogenesis, presents no end of fascinating problems for 

 the embryologist and geneticist, but concerning which there is at present entire 

 lack of knowledge. 



THE GAMETIC AGENTS. 



Since we must admit the existence of specific gametic agents, or factors of 

 composition, and since there is entire probability that the total sum of char- 

 acteristics of the organisms are conditioned by the factors of composition 

 arranged in a system, what about the agents themselves? Are they discrete, 

 invariable imits, present or absent in their totality, or do these agents represent 

 centers of activity, foci, of greater or less magnitude ? The Weismannian con- 

 cept of fixed units was adopted bodily by the neo-Mendelians, and much the 

 same idea is retained by De Vries in his mutation theory; but the fixity of these 

 units seems to be a constant cause of trouble to Mendelian workers ; hence the 

 concept of multiple allelomorphs and kindred concepts developed by Bateson 

 and others, all with the purpose of retaining the concept of the fixed, quantita- 

 tive unit factor as a static morphological element of structure. 



All of our ideas upon this point are based upon our perceptual experiences 

 with the soma ; certain body-characters appear or are absent in their totality ia 

 a given cross ; it follows, therefore, that each is due to a factor which we sym- 

 bolize as A or a, and unless further investigated each is asserted to be a unit 

 factor, an entity. In some other combination, A is shown to break up into two 

 or more portions, and it is therefore a compound allelomorph ; but where is the 

 limit of this compounding? In the first experience the action was a unit or 

 center of activity, while in the second experience a hitherto single center of 

 action, divided into two, each acting independently as far as the end-results are 

 concerned; but what evidence is there from these experiences that we have 

 reached the limit of divisibility ? Further, what logical basis is there in this 

 perceptual experience for the assertion that individualized entities have been 

 discovered or isolated in the operations, or shifted therein ? Conclusions as to 

 the unit nature of the agents in the gametic material productive of these 

 observed results in the soma are based upon the a priori assumption that organ- 



