1904.] NATURAL SCIENCES OF PHILADELPHIA. 745 



corresponds with the ventro-posterior end of the larva; and the end 

 toward which the mesentoderm turns (figs. 23, 24, 26) is anterior; so 

 we can now distinguish anterior and posterior, dorsal and ventral, right 

 and left. The invaginated layer of cells is a mesentoderm, because from 

 it comes both the mcscnchym and the entoderm, as shown in figs. 25-27, 

 where the mesenchym cells are shaded. From that area where the 

 mesentoderm joins the ectoderm (the outer cell-layer) takes place a 

 proliferation of cells of the mesentoderm, leading to the formation of 

 loose cells and masses of cells, the mesenchym, lying within the cleavage- 

 cavity. Fig. 23 shows just at the blastopore two particularly large 

 blastomeres, which I first took to be mesenchym pole cells; but I 

 found them in only one case, and do not believe there is a proliferation 

 from a pair of pole cells, though the proliferation comes from the region 

 near the blastopore. In no case was there any evidence of formation 

 of the mesenchym from the side of the ectoderm; in these stages the 

 mitotic spindles of the ectoderm cells lie always parallel to their free 

 surface, never perpendicular to it, as would be necessarv for the forma- 

 tion of mesenchym. Particularly decisive in this matter are cases like 

 that shown in fig. 25 — mesentoderm cells dividing at right angles to 

 the surface of the mesentoderm. This embryonic tissue is properly 

 ftiesenchym and not mesoderm, since the cells do not arise as an epi- 

 thelium or in compact masses, but separately; and cases like that of 

 fig, 27, where they appear for a time to compose solid masses, are delu- 

 sive and due to the difficulty of distinguishing the cells. At no stage 

 up through the larva do these cellular elements become epithelia. 



The remainder of the early development may be traced rapidly, and 

 simply in order to clarify the organization of the larva. The blastopore 

 closes early {BL, figs. 24-29), or remains as a very narrow opening. 

 The entoderm grows foi-w- ard and is largest at its apex (fig. 26) ; this 

 portion of the cntodermal canal will become the gland of the larva 

 (figs. 2S-30, GL). By cell division the few first-produced mesen- 

 chym cells form more numerous smaller elements within the 

 cleavage-cavity, shown as the shaded cells in figs. 26-32, The 

 next notable change is that at a point nearl}^ opposite the blasto- 

 pore (BL) the ectoderm thickens (fig. 28), becoming two to three 

 cell-layers deep, whereas elsewhere it has become much thinner; at 

 this stage also the anterior end of the cntodermal canal is becoming 

 constricted off from the remainder. In the next stage the thickened 

 ectoderm mass commences to invaginate (fig. 29), this being the com- 

 mencement of the proboscis; at the inner surface (that surface border- 

 ing on the cleavage-ca^■ity) this invaginated ectoderm shows an annular 

 48 



