164 The Mechanism of Bvolutiok in Leptinotaesa 



easy to derive from the parent species, are perfectly homozygous in action, but 

 must be tested by iatraspecific crossings to determine the presence of non- 

 manifested gametic contaminations. These when found must be eliminated. 

 Two such lines reared under the same conditions of the medium for several 

 generations when crossed give a relatively certain test of the reaction proper to 

 the two gametic systems, independent of the conditions of the environment, and 

 not complicated by nonhomogeneity in the stocks. 



In the crossing of these two species I have used only pedigreed materials that 

 had been reduced in cultures to a biotypic condition, and my crosses have been 

 between the decemlineata stock from Chicago (No. 99) and multitceniata from 

 Chapultepec (No. 543), Puebla (No. 541), and Texcoco (No. 547) in Mexico. 

 Of this species I have used the pure multitcBniata biotype 7 and muUilineata bio- 

 type 13. All of these have been of the typical yellow hypodermal color, so that 

 the crosses have not been complicated by the array of hypodermal colorations 

 foimd in the previous cross. The two species are also not greatly different in 

 form-index, except for the narrow form that is present in the biotype muUi- 

 lineata and the broad index that is associated with tacubaycensis. 



After these two species have been reared in the same conditions in the labora- 

 tory at Chicago for three to five generations and then crossed, the uniform result 

 is the dominance of decemlineata in F^ in the juvenile stages, especially in larval 

 body-color, the patterns and larval-pattern sequences being the same in both. 

 The dominance of the decemlineata in P^ can in many respects be altered, by 

 the use of desiccation, high temperatures, and other agents, to diminish the 

 intensity of the F^ red, but in no instance has the color been modified to more 

 than an intermediate condition of yellowish red. The F^ adults show rather 

 uniformly the equal blending of the adult characters of the two parents when 

 the parent lines are in both biotype 7 in pronotum pattern. The pronotel 

 hypodermal color may be variable, swinging between the two parental types, 

 but in the main it is intermediate between them, while the elytral colors are not 

 sufficiently different to produce any change. The form-index, however, shows 

 complete dominance of the decemlineata type, with relatively little oscillation 

 and decemlineata is the dominant type. 



Inbreeding these F^ for the next generation shows that on the whole the two 

 have been through the cross with no dissociation of the characters, there appear- 

 ing in the adults in Fj three classes that are by inspection clearly decemlineata, 

 heterozygotes, and multitceniata. Some fraternities show this separation with 

 remarkable distinctness, others show the presence of intergrading conditions 

 in the heterozygous series, so that the separation of the entire array can not be 

 made by either inspection or measurement, but in such arrays no difficulty is 

 experienced in selecting by inspection the extreme conditions that always breed 

 as extractives, or of the centrally placed heterozygotes. Other conditions in the 

 array are not so easy of calibration and often are only properly placed after test 

 by breeding. This difference in Fj in different fraternities is a product of the 

 conditions surrounding the cross, which in general seem to produce, when con- 

 stant and with little or no oscillation, sharply marked Fj arrangements; but 

 when the conditions are oscillating the groups are less clearly defined or often 

 obscured entirely by the range in the heterozygous members of Fj. 



In these reactions it is certain that the characters of the adult have not been 

 dissociated in passing through the Fi reactions, coming out of the operation in 



