Eeactions and Pkoducts in Intekspecific Ckosses 137 



In crossing of these species it has been the uniform custom to mate in the 

 original crosses only individuals of like condition with regard to this repro- 

 ductive cycle; otherwise troublesome complications in the breeding-out of the 

 series is sure to be encountered. Thus in practice it has been most common to 

 mate the materials on emergence from the resting-period, as one is thus sure to 

 get the first two generations of the series within a half year or less, and thus 

 facilitate the operations. More rarely are matings made between the summer 

 generations which give only F^ before a period of rest intervenes, and only for 

 special purposes are the two aspects of the cycle mated, when the result is 

 complications in abundance. 



These crosses between L. signaticollis and L. undecimlineata have been, as far 

 as this portion of the investigation goes, made entirely between the stocks on 

 emergence from hibernation, and with respect to the character of the reactions 

 the two types of reaction described present quite different results. In the 

 normal cross, when the Ac values are about the same and the conditions of the 

 medium are neutral, the result is the regular cycle in the reproduction of the 

 series, the second generation following at once after the production of F^, giving 

 F, within the usual period of 4 months from the initial cross, and then Y^ is 

 followed by a period of rest in the F3 fraternities, with a duration on the average 

 of from 6 to 10 weeks, depending entirely upon the conditions in the medium. 

 In the second type of result in the crossing of these two species, as shown in 

 plate 14, the result is different. The suspected undecimlineata race, now known 

 to be heterozygous, acts at once as a winter generation, commonly enters within 

 a few days into typical hibernation, and is not to be forced into reproduction, 

 although it may be forced from hibernation. The F^ heterozygous types that 

 are normal give the characteristic response by breeding at once, giving the Fg 

 by the end of 4 months as in the nonnal cross. Further breeding of the two F^ 

 types .show that in the manifestly heterozygous series the reactions in the cycle 

 are normal and can be carried through to at least Fg within 12 or 13 months, 

 while the undecimlineata series in this same time produces only one generation. 



It appears that in this suppressed heterozygous race there are some important 

 rearrangements produced in many of the properties present, and that there have 

 arisen associations of the agents in the gametes, such that the arrangement 

 present allows of only certain visible end-results. There is no doubt that the 

 race is heterozygous, as is demonstrated by the fact that the Fg fraternities show 

 accurate separation of the series into a trimodal polygon, whose significance 

 with regard to the form-index is known; that certain methods permit of the 

 recovery from the race of the nonvisible signaticollis larval and adult elytral 

 characters ; and further, the race shows a complete suppression of the summer- 

 brood production, so that the race appears in a long monobrood cycle that in 

 general aspect is undecimlineata, the total complex being retained without 

 change as far as has been carried and without regard to external conditions as 

 far as is shown in the observations. Only special operations, under precise 

 complexes of the medium, produce dissociation of this complex that arises so 

 uniformly in F^. 



One series of this strain I carried from 1906 to the end of 1914, in which I 



had 10 generations after the initial one, but at the end it was the same in 



behavior as in composition as at the start. One point that came out in this series 



was of much interest. It has been shown that the F, fraternities are trimodal 



10 



