264 The Mechanism of Evolution in Leptinotaesa 



only by the elimination of errors of this kind, and of many others, that an ade- 

 quate test of quantitative accumulation effects can be made. Upon a gross, unre- 

 fined population, or unsimplified strain, no such attempt can profitably be made, 

 and any result obtained will be only a change induced by an unknown or at best 

 by a series of two or more interacting forces. Any permanent change, as far as 

 evidence goes, is most probably due to a new gametic state brought about unin- 

 tentionally by synthetic combination and not to the action of a " selective 

 process." 



EFFECTS OF SYNTHETIC COMBINATION OF DIRECTIONS OF VARIATION 

 IN ELEMENTS OF PATTERN. 



Quantitative modification has, in every instance thus far investigated, resulted 

 in no change of a gametically homogeneous group, clone, genotype, or biotype, 

 but on the other hand, the effects of intelligent synthetic combinations have 

 given different results. This is due to the fact that in these biotypes a system of 

 elements exists in a pattern system, and between these elements many combina- 

 tions are possible, but in every instance it requires the concerted interaction of 

 two or more agents to produce a change ; that is, one element alone, however 

 changed, can not combine with some other gametically ; but it is the experience 

 that both must be involved and show directions of variation towards each other 

 which unite and hold fast the two elements until disrupted by some agency ex- 

 ternal to the particular group. The existence of these bonds or directions of 

 combination affmities differ in their capacity for permanency in the same type 

 and in different types and species. It has already been shown how these bonds 

 change direction ; that their combinations are permanent or transient, and here 

 it remains to show how the biotype groups of the pronotal pattern that can be 

 isolated are also capable of being transmuted by proper treatment. The method 

 of change employed is, in its essential character, organic metathesis, producing 

 interchange and recombination of the characters of the elements of the pattern 

 of the pronotum. 



An examination of the pattern of this part reveals the fact that the h' — a' — 

 a — h group is either widely divergent anteriorly in the form of a widely open V, 

 or parallel with the &' & at the level of the a' a area. This results entirely from 

 the presence of two form-determiners interacting with a pronotal form-factor. 

 I have shown that in the natural material of L. multitceniata there are two gen- 

 eral body-shapes which are expressed by the phenotype group names, " multi- 

 lineata " and " muUitcBniata" Whenever the multitceniata factor is present, 

 the spots h' — a' — a—b are a broadly open V, and when the muUilineata factor is 

 present they are closed and parallel. This can be shown as follows : 



If a perfectly pure homozygous-acting strain of the form melanothorax be 

 obtained, and this combined with the proper forms, tests to prove this above 

 assertion can be made. The melanothorax form is another complex ; the centers 

 are lost, its ontogeny and everything show it to be different from the rest of the 

 species when in pure strains, even though it is not to be distinguished in the adult 

 condition. If such a pure strain is crossed with a biotype 12, also homozygous in 

 action, there results in F^ a progeny with the melanothorax dominant, either 

 completely or nearly so. These inbred give in Fg three complex groups — one 

 with the form of multitceniata, one intermediate, and one of the muUilineata 

 form. The proportions are variable, but in the long run are a close approximate 



