138 The Mechanism op Eyoltjtion in Leptinotaksa 



in trihybrids, entirely comparable with those investigated by others, and in aU 

 respects a duplicate of the classic trihybrid cross of Mendel with peas. The 

 contrasted characters represent groups of agents associated in reaction that are 

 transferred entire in the rearrangements of the qualities in F^. Two points are 

 of especial interest in this series, namely, that the characters are successive in 

 ontogenetic sequence, a condition quite different from that present ia most other 

 trihybrid crosses examined, and the presence in the series of a meristic series 

 of color-marks that is iaherited as a unit system. 



I have repeatedly made this cross in both directions, testing to Fj or beyond, 

 as far as space and aid permitted, and in all there has been not one indication 

 that the behavior in any respect was other than the most orderly and typical. 

 This statement applies only to the cross when made under conditions that are 

 neutral to the two, and when the Ac determiner has the value of about Ac'° in 

 both, so that the reaction represented with this complex, and conditions in the 

 medium, is the basal reaction between the two species complexes. The same 

 cross has given most perplexing arrays at different times, the production of 

 which are now to some degree understood. It is important to make certaia that 

 the reaction between the gametic complexes is typical, or else to determine the 

 agents that produce the discordant result. In table 15 I have given examples of 

 data derived in the testing of this cross in Fj, where the Ac values were about 

 60 and the conditions in the medium were neutral. 



One peculiarity present in this series is the dominance in Fj of the larval 

 color-characters of the female parent in the cross, unless the conditions are in 

 one way or another inhibitory to this, a change which is easy to produce experi- 

 mentally. There is no indication that the dominance is in any manner asso- 

 ciated with sex, or any other sex-relation. The corresponding groups of deter- 

 miaers are introduced by the male gamete and are present in full intensity in Fj, 

 so that there is no suggestion of the contrasted characters being absent, and the 

 fact that the same result occur^ in the reciprocal at once throws out any idea of 

 the strength of one parent species being superior to the other. The fact that 

 conditions ia the medium were able to easily change the relations of the four 

 groups of agents in the larval stages, as far as their manifestation in Fj is con- 

 cerned, but not in any way in the Fj array, leads to the conclusion that the con- 

 dition of apparent female dominance in the Fj juvenile stages represents a 

 condition ia the mass of the female gamete, already has an initial velocity along 

 the line from which it came, that is imparted to it by the developmental activi- 

 ties which produced the matured and which egg determines F^ juvenile domi- 

 nance. Therefore, the initial developmental momentum in the egg continues 

 ia the same direction these groups of agents as uppermost in visibility, until 

 they are ia one way or another, by internal or external conditions, retarded until 

 the two sets of agents present are on a level, and then the visible dominance in 

 the Fi heterozygotes depends entirely upon the role of conditions within the 

 organism and without it, a relation that is shown by simple experiments, so that 

 the reciprocal F^ juvenile arrays can, within the limits of the time and change 

 possible with the different molts, be altered at will. This is especially true with 

 regard to the lipoid body-colors, which can be changed about at vdll ia the larvae 

 merely by the alteration of conditions. I am of the opinion that this is the 

 cause of the conditions found in the Fj crosses of Fundulus, described by New- 



