42 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 98 



derivation, independent of the mesoderm, is not new to the literature 

 of insect embryology. In 1897 Carriere and Burger described such a 

 method of entoderm origination in both Chalk odoma and Tenebrio. 

 Noack (1901) says that in Calliphora the entoderm cells are produced 

 in the same manner. It is of interest to note that Uzel (1897a, 1898) 

 states that the entoderm of the apterygote forms Lepisma and 

 Campodca is likewise formed from cells which migrate in from the 

 blastoderm. Philiptschenko (1912) describes the same origin for 

 I sot o ma. However, in these last two cases, in addition to the anterior 

 and posterior rudiments, entoderm is proliferated along the embryonic 

 band. 



EXTERNAL EVIDENCES OF SEGMENTATION AND CHANGES 

 IN EMBRYONIC SHAPE 



Very early in the development of flea embryos, at the beginning of 

 the second day and even before the differentiation of the germ layers, 

 there becomes evident what appears to be a precocious segmentation 

 of the ventral plate. The manifestations of this pseudosegmentation 

 take the form of incomplete transverse zones of the ventral blasto- 

 derm. They are incomplete in that the middle portion of the plate 

 (consisting of the cells of the future mesoderm) is unaffected by 

 this phenomenon. At first it appears that these precocious divisions 

 correspond to the primary segmentation of the insect germ band which 

 was first observed by Ayers (1884) in Occanthus and later by Graber 

 (1888a, 1890) in a number of different insects. These authors were 

 able to distinguish four general regions in the early germ band. They 

 were a primary cephalic, a maxillary, a thoracic, and an abdominal 

 region. Graber attributed great phylogenetic significance to these 

 divisions which he regarded as the primary body segments. Ac- 

 cordingly, he termed them macrosomites in contrast to the definitive 

 body segments or microsomites into which they subsequently divide. 



Careful examination of flea embryos indicates that these general 

 zones are more numerous than the four regions observed by Ayers 

 and Graber. In fact their number is not constant and what is more, 

 they soon disappear. Therefore, they are to be regarded as pseudo- 

 segments which are due to some other factor than segmentation, 

 perhaps a mechanical contraction. Wheeler (1889) describes a some- 

 what similar phenomenon in Lcptiiwtarsa and ascribes it to a 

 wrinkling of the ventral plate. The fact that such segmental appear- 

 ance is not evident in longitudinal sections of flea eggs of this stage 

 substantiates this explanation. 



