NO. 3 EMBRYOLOGY OF FLEAS KESSEL 35 



derm, excepting the germ cells, have been columnar in shape. At 

 this stage, however, the cells which are to form the anterior mesenteron 

 rudiment assume an irregularly cuboidal or spherical outline. A simi- 

 lar appearance of the cells comprising the mesenteron rudiments of 

 Leptinotarsa (Wheeler, 1889) and Apis (Nelson, 191 5) has been 

 observed. 



The mesenteron rudiments of fleas are not originated by a process 

 of cellular proliferation from the blastoderm such as Carriere and 

 Burger (1897) have described for Chalicodoma. Neither is there an 

 invagination of the blastodermic layer in its formation. Instead, the 

 process is by migration of cells from the ventral blastoderm. With 

 the changes in the appearance of the future mesenteron rudiment cells, 

 which have been described above, the cells themselves tend to become 

 loosely arranged. This loosening results in the appearance of numer- 

 ous minute interstices between the cells of the rudiment anlage. It is 

 probably responsible for the rounding out of the cells as well. The 

 interstices are not apparent for long, as the cells soon migrate below 

 the surface in an apparently confused mass and then become compactly 

 arranged to form the rudiment proper. The loss incurred by the 

 emigration of cells from the surface is gradually made up by the 

 approximation of the blastoderm over the mesenteron rudiment. For 

 some time previous to this, however, there remains an irregularly out- 

 lined and extremely shallow pit (pi. 6, fig. 53) over the mass of cells 

 which may now be regarded as constituting the entoderm. The process 

 up to this time is similar to that found in Apis (Nelson, 1915). In 

 flea eggs, however, the shallow pit is finally closed over by the blasto- 

 derm (now ectoderm at this point on the surface of the egg mass) 

 in contrast to the case of Apis, in the development of which Nelson 

 observed that a plug of mesenteron cells reaching to the external 

 surface remained after the approximation. In spite of the fact that 

 it completely covers the rudiment of the mesenteron, the ectodermal 

 sheet over this area in flea ova may be slightly indented to form a 

 second shallow depression which is to be recognized as the location 

 for the stomodaeal invagination. 



While there is no evidence that cell proliferation has played any 

 part in the formation of the anterior mesenteron rudiment up to this 

 time, subsequently the rudiment increases in thickness by the multi- 

 plication of its cells. This is shown by the presence of mitotic figures 

 among the cells of the lower portion of the entoderm mass. The 

 anterior margin of this anterior mesenteron rudiment is thinned out 

 rather abruptly and terminates at a point which is located slightly 



