164 



TEXT-BOOK OF ENTOMOLOGY 



on (p. 921) he concludes that in the Hymenoptera the "fifth " segment (first ab- 

 dominal) is not in reality a part of the true thorax, " but is sometimes connected 

 more or less with that region, or with the abdomen, being intermediate between 

 the two. Hence we have ventured to designate it the thoracico-abdominal seg- 

 ment." Had he considered the higher Hymenoptera alone, he would undoubtedly 

 have adopted Latreille's view, but he saw that in the saw-flies and Lepidoptera 

 the first abdominal segment is not entirely united with the thorax, being still 

 connected with the abdomen as well as the thorax. Reinhard in 1865 reaffirmed 



Latreille's view. In 1866 we stated from observa- 

 tions on the larvae made three years earlier, that 

 during the semipupa stage of Bombus the entire 

 first abdominal segment is "transferred from the 

 abdomen to the thorax with which it is intimately 

 united in the Hymenoptera," and we added that 

 we deemed this to be " the most essential zoo- 

 logical character separating the Hymenoptera from 

 all other insects." (See Fig. 93, showing the 

 gradual transfer and fusion of this segment with 

 the thorax.) In the saw-flies the fusion is incom- 

 plete, as also in the Lepidoptera, while in the 

 Diptera and all other orders the thorax consists 

 of but three segments. (See also pp. 90-92.) 



The cercopoda. We have applied this 

 name to the pair of anal cerci appended 

 to the tenth abdominal segment, and which 

 are generally regarded as true abdominal 

 legs. As is now well known, the embryos 

 of insects of different orders have numerous 

 temporary pairs of abdominal appendages 

 which arise in the same manner, have the 

 same embryonic structure, and are placed in 

 a position homologous with those of the 

 thorax. In the embryo of (Ecanthus rudi- 

 mentary legs appear, as shown by Ayers, 

 on the first to tenth abdominal segment, 

 the last or tenth pair becoming the cerco- 

 poda; and similar rudimentaiy appendages 

 have been detected in the embryos of 

 Coleoptera, Lepidoptera, and Hymenoptera 

 (Apidse). Cholodkowsky has observed 

 eleven pairs of abdominal appendages in Phyllodromia. 



They are very long and multiarticulate in the Thy sanura (Fig. 

 179). In the Dermaptera they are not jointed and are forcep-like. 

 It should also be observed that in the larva or Sisyra (Fig. 181) 

 there are seven pairs of 5-jointed abdominal appendages, though 

 these may be secondary, structures or tracheal gills. In the Per- 



FIG. 179. Abdomen of Ma- 

 rlnUx iiKiritiiini. 9 seen from 

 beneath : the left half of the 8th 

 ventral plate removed; I-IX, 

 abdominal segments; c, cercopoda; 

 <!>, coxal glands ; hs, coxal stylets ; 

 If, ovipositor. After Oudemans, 

 from Lang. 



