103 The Mechanism of Etolutiok in Leptinotaesa 



gametic complexes, between which characters might be shifted, producing in the 

 end a considerable array of types in F^ and F3 differing in small but none the 

 less distinctive and permanent aspects. No such metathesis is observed in thia 

 series, and I believe that this type of reaction represents the simplest that can 

 exist in the crossing of species. 



It may be objected that the materials are not species because they cross and 

 give fertile Fj and Fj progeny, and the fact that they cross at all would indicate 

 to many a lack of specific distinctness. Whether the above objection is tenable 

 or not need not be discussed, as it is simply a matter of anticipatory definition. 

 When the species come into the laboratory from nature they are less fertile and 

 often cross with diflSculty, producing quite different results, as will be shown 

 presently. They are reared and crossed under the conditions of experiment 

 given, that is, under the mean average of each chief environmental factor of their 

 breediag-period ia nature, so that the conditions of experiment may not intro- 

 duce any incidental agents that might change them; only the rhythm of native 

 habitats is changed, and instead of the pressure of the brief season in the habitat 

 of L. signaticollis, both are given continuous mean breeding-season conditions, 

 to which the species adjust themselves without discoverable change except in the 

 rates of reaction in development. As the two become attuned to the same con- 

 ditions in the medium, their interfertility increases until in the end the fertility 

 is nearly as high in interspecific crosses as it is in pure-line matings. When thia 

 state is reached the reactions shown in the crossing of these two speciea are 

 uniform and represent the true basic reaction of the two gametic systems when 

 combined in crossing under neutral conditions of the medium, and it is the 

 reaction of the systems in simplest terms under conditions that permit of the 

 manifestation being entirely the product of the composition and structure of the 

 interacting masses. 



Crosses of the kind indicated between these two species have been repeated 

 yearly since 1907 as part of the annual routine to test this point and also the 

 condition of the stocks, with no indication of any differences whatsoever. These 

 yearly tests are random matings, with random testing into Fj, and sometimea 

 into Fg or F^, depending upon the space and help available. 



MODIFICATIONS OF THE NORMAL REACTION. 



If the crossing of these two species presented only the behavior shown, it 

 would be of itself of interest because of its simplicity and the integrity of the 

 gametic systems shown, but this is not the reaction shown when the materials 

 are fresh from nature and not adjusted in their reactions to the same set of 

 external conditions. I found in 1904 that these two species, when fresh from 

 nature, in guarded parentages, gave in F^ a behavior that suggested that one of 

 the parents was heterozygous, the other homozygous with respect to the elytral 

 striping, with the result that in Fj two classes of about equal numbers appeared. 

 Thus when the cross was made between L. signaticollis female and male 

 L. diversa, the uniform result in Fj was about 50 per cent of mid-types and 

 50 per cent of the female signaticollis types, and these when inbred in each class 

 gave in F^ and in subsequent generations the extracted female signaticollis type, 

 "breeding true" and the mid-type splitting in Fj into signaticollis, mid-, and 

 diversa types in the usual ratios. This result I have shown in plate 9. It is in 

 every way typical and suggestive of the crossing of a DD X DR or DR x RR. 



