158 The Mechanism of Evolution in Leptinotaksa 



dissociated and recombined in the crossing of these two species. In this race, 

 the subject of the test, there is clearly present in the interactions two groups of 

 agents, or two agents, a form-factor group that is not dissociated and a form- 

 determiner, the iateraction of the two being to produce the characteristic form 

 or type as measured by the form-index, or recognized by inspection in many 

 instances in its pattern modifications. These two groups of agents are purely 

 symbolizations at present and no indication is to be had that shows what they are. 



The race tested is in all respects decemlineata, with the exception of form, 

 which is that of oblongata, while its breeding true shows that there has been 

 formed a permanent combination, that is, a harmoniously acting one, so that it 

 gives uniform fraternal arrays, generation after generation, the usual test of 

 purity and stability. That the race is due to the combination of the decem- 

 lineata complex with the form determiner from oblongata is shown in the test 

 cross with melanothorax in which, in the complex interchange that takes place, 

 there comes out a decemlineata form as the result of the combination of the 

 decemlineata form-factor complex with the determiner from the melanothorax 

 race. This result could only have come about in two ways ; either it is the prod- 

 uct of the process stated, or the production of the decemlineata form in Fj is 

 due to the presence in the tested line of the decemlineata determiner being pres- 

 ent in hidden form in the tested race and not acting. However, tests in cross- 

 ing with the pure decemlineata and oblongata have failed to show the existence 

 of it in reactions where it would have been revealed if present. It is hardly to 

 be expected that the race that is being tested is due to the transfer of its very 

 numerous decemlineata characters to an oblongata base. There are too many 

 of these and too decided an absence of oblongata attributes in the combination, 

 nor is the combination precisely the oblongata parent condition, but below it, 

 towards decemlineata, showing that the combination does not react in the same 

 manner as in either of the parental stocks. The fact that the line is in color, 

 reaction, odor, habits, development, and in all respects decemlineata, save in the 

 element of form which is nearly that of oblongata, the experiences in the tests 

 of the line that have been made force one to the conclusion that there has been 

 produced a combination of agents from oblongata with the decemlineata com- 

 plex, producing a new, homozygously acting genetic line. The series shows, I 

 believe, the dissociation and recombination of the form-producing agents in the 

 gamete and the existence of two groups, the interaction of which is necessary to 

 the production of the specific end-result, a form-factor group, and a determiner 

 group, and the evidence shows that it is the determiner group that is transferred 

 and recombined and not the factor series. It is further shown in the test made 

 with melanothorax that the pronotal-pattem series are capable of interchange 

 as a whole, another example of the action as a whole in crossing of meristic series 

 of characters, other examples being the behavior of the larval patterns in this as 

 in other series of crosses already described. 



Three matings of the (a) type in P^ have given different results from those 

 presented, and have not shown the presence in Fj and in later generations or in 

 tests of the breaking up of the line into the different form-types shown in the 

 series just described. These three lines have been uniformly decemlineata in 

 form-index and showed only the production of diverse ¥^ groupings dependent 

 upon the metathesis of different agents that came from the parent stocks, espe- 



