NO. 3 EMBRYOLOGY OF FLEAS KESSEL 3 



it must be recognized that much of the past work in the field of insect 

 embryology is inaccurate. This is evidenced by the distinctly opposite 

 interpretations which frequently have been given by workers on the ■ 

 same species. In view of this fact it is inadvisable for the writer to 

 attempt a phylogenetic application of the present observations until 

 he has personally investigated the embryology of those forms which 

 are suspected of being most closely related to the Siphonaptera. Only 

 then can it be certain that the same methods and interpretations have 

 been applied to the different subjects and that the comparisons are 

 worthwhile. 



SPECIES 



The observations made during this study indicate that the embryo- 

 logical development of all fleas is essentially the same. Except for 

 such superficial characteristics as size, nature of the chorion, and the 

 number of micropyles, the corresponding stages of the eggs of dif- 

 ferent species appear to be identical. This was shown by making 

 careful studies on the development of three species of diverse system- 

 atic positions. These included Ctenocephalides jells (Bouche), the cat 

 flea; Nosopsyllus jasciatiis (Bosc), the common rat flea of temperate 

 regions ; and Hystrichopsylla dippiei Roths., a giant form which lives 

 in the nests of the wood rat Neotoma. 



Following the classification of Oudemans (1909), the order Siphon- 

 aptera is divided into two suborders on the basis of the presence or 

 absence of a dorsal suture on the head. Fleas which possess such an 

 incrassation are placed in the suborder Fracticipita, while those which 

 lack it constitute the suborder Integricipita. If this division is a 

 natural one based on true relationships, the choice of the three species 

 used in this study is valid. Ctenocephalides jells and Nosopsyllus 

 jasciatiis belong to the Integricipita, while Hystrichopsylla dippiei is 

 classed in the Fracticipita. 



In contrast to Oudemans' system, many authors, including Ewing 

 (1929), consider that the subdivision of the Siphonaptera into these 

 two groups is artificial and unwarranted owing to the number of 

 intermediate forms in which it is difficult to determine whether or not 

 a dorsal suture is present, and also because of the lack of sufficient 

 correlating characters. These authors prefer to follow the classifica- 

 tion of Baker (1905) which regards the Siphonaptera as a compact 

 group and, accordingly, divides it directly into families. Based on 

 this classification, the choice of species for this study is again found 

 to be a satisfactory one inasmuch as the three forms selected belong 

 to three different families. Ctenocephalides jelis represents the Puli- 



