270 C. O. WHITMAN. 



in water to near the boiling point and then treated with alcohol, 

 the mesoblasts can easily be removed with needles without 

 breaking. In this way I have generally found a nipple-like 

 protuberance at the upper pole of each, which shows that they 

 are in process of proliferation. 



I have not succeeded in obtaining sections in which the full 

 amphiastral division could be seen ; but I have preparations 

 which show parts of the amphiasters and the connection of the 

 same with the nipple-like protuberances. This proliferation 

 begins as early as stage 33 (24 h. 15 rain.), at least with the 

 left mesoblast. The path of cells leading Irom the upper pole 

 of this mesoblast may easily be seen on the living egg in this 

 and some of the following stages (figs. 33 — 38). 



{g) Relation of the Mesoblasts and Neuroblasts to the Germ- 

 bands. — As before remarked, the cells of the blastodisc multiply 

 first, at the expense of the primary segments {a, h, c). In stage 

 33, and perhaps earlier, the two mesoblasts begni to produce cells 

 that take their ])laces below the lateral edges of the blastodisc 

 (figs. 81, 82, 84^, 85, 87). Those produced by x ^, although at 

 first visible from the surface, are soon covered by the smaller 

 cells of the centrifugally expanding disc. The eight neuroblasts 

 begin as early as stage 37 to take a conspicuous part in the cell- 

 proliferation. From this time (£8h. 30m.) the germinal disc 

 receives new material from only two ditferent sources, namely, 

 the neuroblasts and the mesoblasts. It is at this epoch that we 

 begin to see distinctly lines of cells leading away from the inner 

 nuclear poles of the neuroblasts. In stage 38 (36h.) the lateral 

 borders of the disc are plainly thickened and transversely arched. 

 These thickened borders are the germ-bands (nota3 primitivjc). 

 Each of these " embryoplastic " bands is composed of four 

 longitudinal lines of cells produced by the neuroblasts, and of 

 larger subjacent cells produced by one of the mesoblasts. The 

 linear arrangement of the neuroblast-products may be seen on 

 the living egg, but is more distinct after treatment with osmic 

 or chromic acid. The two bands taper a little towards the 

 cephalic ends, which terminate near the boundary lines of the 

 ventral blastomere (c.) 



The signification of the neuroblasts has not been hitherto 

 understood. 



Grube ( ivliVr) found only two — at most three — at the anal 

 end of each band. He sujjposed that they contributed elements 

 to these bands, but to what extent and to what purpose was 

 unknown. Rathke (,^-^,3^*—^ found three joined by as many 

 rows of cells to the posterior end of each baud, and was inclined 

 to believe that they entered into the structure of the terminal 



