124 The Mechanism of Evolution in Leptinotaksa 



This instance is, I believe, the simplest thus far described in the study of mono- 

 h3^brid crosses and represents the simplest type of operation that can at present 

 be thought of in the intercrossing of two natural species. The only complexity 

 that enters into the series is that due to the differing values oi Ac and the role 

 of conditions in the climatic complex in inducing changes in the gametic 

 system through the activity of Ac. 



The nature of fixed hybrids is suggestively indicated, as are some of the 

 alleged "permanent blends" between species. So, too, it is strongly suggestive 

 of the manner of production of many of the sports or mutants that occur, 

 especially in the crossing of domesticated races, and it may well be true for the 

 species in nature. Whether this is in any way related to the mutation behavior 

 found by De Vries in Oenothera will receive attention in its proper place. 



LEPTINOTARSA UNDECIMLINEATA X LEPTINOTARSA SIGNATICOLLIS. 



Crossing of these two species in its simplest conditions always shows reactions 

 of a trihybrid nature, and is, therefore, more complex in its relations and reac- 

 tions than the crosses of signaticoHisxdiversa. These two species differ in 

 many respects in the adult, and in the juvenile stages inhabit entirely unlike 

 environmental complexes, with consequent differences in many of their 

 reactions. 



In the adult condition there is difference in the form index, the elytra have 

 the same difference as in the diversa X signaticolKs cross, undecimlineata 

 being striped, signaticollis not, due to the same set of agents present in the 

 gametes. These agents, while similar, are not in diversa and undecimlineata 

 the same, and do not produce the same end-result ; that is, the elytral complex 

 from diversa is alternative to that from undecimlineata when the two are 

 crossed, the differences showing in the results of the 'Pu determiner and in the 

 character of the stripes produced by the V determiner. 



In the juvenile stages there are differences of lipoid color and in the pattern 

 of the second and third larval stages. In undecimlineata the body-color is 

 white, with only the spiracular spots present in all of the stages ; in signaticollis 

 it is yellow, with only the spiracular spots present in the second and the full 

 system of spots in the third stage. Three alternative pairs of characters are 

 formed, no one of which is a simple one, namely, the unresolved species residue 

 characterized by the elytral-pattern group, the larval-pattern group, and the 

 lipoid-color group. There are others present, but they do not in this series 

 become dissociated from the general species complex, which, as far as discovered, 

 retains its integrity in the same manner that was present in the cross between 

 signaticollis and diversa. These three groups of agents are "dissociated" in the 

 ordinary operations of crossing. In addition to the agents mentioned, there are 

 also present in this cross the differences in the Ac determiner found in the pre- 

 vious instance, only in this the average value of Ac in L. undecimlineata is, in my 

 cultures, 64 days, same as in nature, within small range. In table 13 are 

 given figures, from cultures of this species, in the laboratory at Chicago and in 

 nature, showing the length of ontogeny, that is, the value of the Ac determiner, 

 which is not changed in this species by introduction into the laboratory. 



The two species live in quite different conditions; undecimlineata living in 

 regions of high humidity and temperature, low evaporation rates, and long 



