CELL-LIXEA GE. 21 



some cases from the third quartet {Physa, Ancia{}))} Assum- 

 ing this to be the case, what shall we say of the mesoblast- 

 bands, which are in annelids and mollusks derived from the 

 fourth quartet and which, as we have seen reason to conclude 

 (p. 12), are probably to be regarded as derivatives of the primi- 

 tive archenteron ? The development of the polyclade suggests 

 an answer to this question which is in harmony with the facts 

 discussed in the first part of this paper. As earlier observers 

 have shown, the fourth division of the *' macromeres " in the 

 polyclade is unequal, giving rise to four smaller cells at the 

 lower pole of the embryo {A-D, Fig. 6, C-E), and to four much 

 larger cells lying above them. From these eight cells, which 

 are heavily laden with deutoplasm and differ entirely in appear- 

 ance from the ectomeres and mesomeres, the archenteron is 

 formed. With this Leptoplana exactly agrees, and I can find 

 no evidence that mesoblast-cells are formed from any of these 

 eight cells. If now we judge solely by relative position without 

 respect to size, the four larger cells or ''macromeres" (4—4) 

 correspond exactly with the fourth quartet of annelids and mol- 

 lusks — in fact, they are relatively not very much larger than in 

 some of the mollusks (e. g., Planorbis, t. Rabl, 1880). Lang 

 discovered the remarkable fact that in Discoco^lis, as in so many 

 of the latter animals, the posterior cell of these four di- 

 vides long before the others ; and further, that this division is 

 equal, giving rise to two symmetrically placed cells at the pos- 

 terior end of the embryo, while the ensuing divisions of the 

 other three cells of the quartet are unequal and irregular.^ 

 Mead^ has pointed out the very remarkable resemblance of 

 these two cells in Discocoelis to the "primary mesoblasts " of 

 annelids and gasteropods and even goes so far as to suggest 

 that they may give rise to mesoblast-bands in the polyclade. 

 My observations on Leptoplana lend no support to this sugges- 

 tion, agreeing nearly with those of Lang on Discoea^lis save in 



^ Edouard Meyer (1890, p. 299) has definitely compared the "parenchyma" 

 (mesoblast) of the Turbellaria with the "larval mesenchyme" of the annelids, 

 which he believes to have a different origin from the mesoblast-bands. 



2Cf. Lang, 1884, Figs. 17-20. 



31897, p. 289. 



