Eeactions and Products in Interspecific Crosses 



125 



seasons for development, longer even than in the habitat of L. diversa, in sharp 

 contrast with the conditions in the habitat complex of L. signaticollis. The 

 first results obtained in the crossing of these two species in 1904 and 1905 were 

 most confusing. 



When these two species are crossed under common conditions, with the values 

 of J. c in each about 60, the uniform result is the production, regardless of the 

 direction of the cross in F^, of a uniform progeny that are intermediate in the 

 adults, with the undecimlineafa larval groups dominant when the female parent 

 is undecimlineata, and with the signaticollis larval series dominant when 

 signaticollis is the female parent. The real dominant, if such there be, is 

 undecimlineata, as shown in the working out of the F, array. In F2 there is 

 produced the typical array of a trihybrid. In plate 13 is shown the results pro- 

 duced when these two species are crossed, when the Ac values are the same and 

 the conditions in the medium neutral, and the same result is produced when the 

 cross is made in the opposite direction, the only difference being the dominance 

 in the juvenile stages of the female parent over the male in the juvenile charac- 

 ter, so that the only difference between the reciprocal crosses is the dominance 

 in Fi of the larval complexes characteristic of the female parent. 



Table 13. 



Note. — The Chicago records are random samples from the breeding stocks at Chicafro, taken from Fi to Fio. 

 The records from nature are from tests of breeding in the native habitat made mostly in 1905 and 1906. 



If in the crossing of these two species the conditions of the medium be those 

 used in the normal cross of signaticollis and diversa, the dominance in F^ is 

 always of the female parent in larval characters, but in the adult the result is 

 always uniformly mid-types strictly intermediate between the parents. The 

 Fi dominance of the female larval characters may be changed partly or com- 

 pletely by the conditions present in the medium, or incident upon the progeny 

 during ontogeny, so that the aspect of the F^ larvae is variable, depending 

 upon the conditions of life. Under neutral conditions F^, while variable in the 

 manifestation of the larval condition, is constant in the adult array and 

 remarkably uniform in the F2 products that come out. 



While none of the groups of agents or characters that are alternative are 

 simple, but are in instances compound associations of agents whose collective 

 interaction is necessary to produce the character, the reaction in this cross is 



