36 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 98 



posterior to the point of origin of the developing amnio-serosal fold. 

 The short interval of undifferentiated hlastoderm which separates the 

 two is some dozen cells in length (pi. 2, fig. 25 ; pi. 6, fig. 54). 



The origin of the posterior mesenteron rudiment, although essen- 

 tially similar to that of its anterior counterpart, differs in certain 

 details. The region of appearance in this case is dorsal to the pole 

 rather than ventral. Furthermore, the length of the blastoderm con- 

 cerned in its formation is considerably greater. The initial manifesta- 

 tion of the formation of the posterior mesenteron rudiment is a thick- 

 ening, accompanied by a flattening of the blastoderm slightly dorsal 

 to the posterior pole of the egg (pi. 2, fig. 14; pi. 6, fig. 51). Owing 

 to the later involution of the posterior portion of the germ band, 

 the anlage of this rudiment is drawn from its place of origin to a 

 point yet farther around the posterior pole of the egg. The greater 

 length of the posterior rudiment, as contrasted to the anterior one, 

 is noticeable at all stages in its development. Furthermore, its margins 

 on all sides are less abrupt. Although the posterior rudiment tends 

 to be somewhat thinner than the anterior rudiment at corresponding 

 stages of development, it ultimately attains an equivalent thickness. 

 This thickening takes place during the later phases of the involution 

 of the germ band and is not completed until the tail of the embryo 

 has reached the center of the egg. Figures 59 and 60 (pi. 7) show this 

 aggregate of mesenteron cells lying immediately within the ectoderm 

 adjacent to the posterior terminus of the amnio-proctodaeal cavity. 

 Although a short portion of the ventral plate is continued around this 

 terminus to connect with the amnion, the cells of the posterior mesen- 

 teron rudiment are not included in this extension. 



The appearance of the two mesenteron rudiments marks the first 

 step in the differentiation of the germ layers in flea development, but 

 there soon follows a change in the nature and thickness of the 

 remaining parts of the ventral plate. This consists of the process 

 usually called gastrulation. In the development of fleas, as of other 

 insects, this process results in the transformation of the blastoderm 

 into a double-layered germ band. The superficial layer in this case is 

 continuous with the sheets which cover the two mesenteron rudiments 

 and like those coverings is to be regarded as ectoderm. The lower 

 layer is continuous with the mesenteron rudiments, there being no 

 sharp line of demarcation between either of them and this lower layer. 

 Because of the part played by this lower layer in the organogenesis 

 of the embryo, it is to be identified as mesoderm. 



The differentiation of the blastoderm of the ventral plate region 

 in that section between the mesenteron rudiments is not identical in 



