408 C. H. TURNER 



Although the organization of termite society resembles in 

 many respects that of the ants and bees, yet their mating 

 behavior is quite different. According to Snyder (99) the sexual 

 relations of termites is continuous. Copulation is repeated at 

 regular intervals for several years. 



The advent of the tropism hypothesis induced students of 

 behavior to look upon the hovering of insects as a tropism, 

 usually an anaemotropism. In 1908 3 it was demonstrated that 

 the hoveling of one species of mining bees is a mating device. 

 Tn 1911 Perez* reported observations which induced him to con- 

 clude that the hoverings of several species of flies are prelimi- 

 nary to mating. Records of the field work of Turner (110A) 

 on the ant Lasius niger and of Rau (82) on the solitary bee 

 Colletes compactus show that the mating of both of these forms 

 is preceded by a riotous sun-dance of the males. Occasionally 

 females appear in the midst of the dancers. Then certain males 

 drop out and mate. Apparently, in many cases, we must look 

 upon the hoverings of insects, not as tropisms, but as prenuptial 

 dances of the males. 



According to C. L. Turner (111): " 1. Movements prelimi- 

 nary to copulation are fairly constant in each group of Orthop- 

 tera and vary from simple (Mantidae, Phasmidae and Acrididae) 

 to complex (Blatidae, Gryllidae, and Locustidae). 2. There is 

 sex discrimination in the males of all forms. The female plays 

 an aggressive part and displays discrimination of sex in some 

 groups while in others she is absolutely passive. 3. There is a 

 typical mode of copulation for each family of the Orthoptera. 

 In the Mantidae, the Phasmidae and the Acrididae there is 

 a superposition of the body of the male. In the Blatidae and 

 Gryllidae there is a superposition of the body of the female. 

 In the Locustidae there is an end to end copulation. 4. Families 

 represented by the least number of sub-families are highly spe- 

 cialized; while those represented by the largest number of sub- 

 families have a generalized type of reproductive behavior. 

 5. A comparison between a classification based upon the repro- 

 ductive behavior and one based upon paleontology cal evidence 

 shows a striking agreement and suggests that the different types 



3 Turner, C. H. The Sun-Dance of Mellissodes. Psyche, 1908, pp. 122-124. 



4 Sur quelques Particularit£s curieuses du Rapprochement des Sexes chez cer- 

 tains Dipteres. Bull. Scientifique de la Belgique, 7th Series, vol. XLV, pp. 1-14. 



