Ebactioks and Peoducts in Intbespecific Ceosses 179 



ate themj are of the nature of what has been called fixed hybrids, niunerous 

 examples of which have been described in the literature, especially that of horti- 

 culture and in the older writings. There would be little question as to the 

 stability of these types and of their position in nature, and no doubt they would 

 receive at the hands of taxonomists the designation of species. However, it is 

 shown in the examples that have come up in the course of these experiments that 

 they are heterozygotes, of such arrangement of the factorial agents that the 

 separation of the two gametic systems in Fj gametogenesis is either obscured as 

 in the fixed heterozygous lines arising from the crossing of signaticolliSj diversa, 

 and undecimlineata, or suppressed, as in the products of the cross of diversa and 

 decemlineata. As far as the evidence shows us this diversity is purely a result 

 of the factorial action, within and without, that is productive of these conditions 

 in the crossing of species. 



An interesting series of data are derived from these species crosses, showing 

 that the Mendelian reaction, which is so universally present in them all, is a type 

 of reaction that is entirely independent of the size of the factors that are trans- 

 posed between the two crossed species. It is interesting to find, as in the cross 

 of diversa and signaticollis, when the Ac values are similar, that the reaction 

 in crossing is in all respects most accurately and typically Mendelian in its 

 showing of Pj and Fa arrays; but more of interest is the action of the entire 

 gametic complex of each of the parent species as a unit in the operations, giving 

 the most perfect monohybrid displays. There is no dissociation from the 

 masses ; each in all experiments retains its integrity ; and the resultant reaction 

 is of the purest monohybrid type possible, and complications come in only when 

 the parent stocks differ in the Ac values, which differences produce not dissoci- 

 ations but derangements in the nature of the Pi heterozygotes, giving arrays that 

 might easily lead to false conclusions, as the nature of the gametic constitution 

 of the parent stocks. 



In more complicated crossings, as in that between signaticollis and undecim- 

 lineata, the results show uniformly the retention of the integrity in the main of 

 each of the gametic systems, but with some dissociation of certain juvenile char- 

 acters of color and pattern that are transposed in Pj gametogenesis, giving the 

 expected chance array of the Pa combinations of different degrees of stability, 

 precisely following theoretical expectation on the basis of the trihybrid Men- 

 delian reaction. Also, there arise the same general type of complications from 

 differences in the Ac values of this cross. Between this cross and the preceding 

 certain important differences can not fail to be noted. 



These two crosses show that the Mendelian reaction in type and in operation 

 is independent of the character of the substances involved, and this is shown 

 clearly by the action in the first cross of the total gametic system as a unit, and 

 in the second by the action of the larger portion of each of the two systems as 

 units and the simultaneous action of two portions of them also as units. These 

 two dissociated portions — the color-determiners and the group of agents produc- 

 tive of the larval pattern— are not of the same nature or extent. One is probably 

 a single agent, not capable of present change or dissociation into smaller agents, 

 and this is the condition of the yellow-color-determiner that is dissociated and 

 recombined. The larval pattern is known to be a system of minor agents that 

 collectively act to produce the pattern, a conclusion that is warranted by the 

 capacity shown for alteration of this pattern and the removal of specific portions 



