26 RECORD OF CURRENT RESEARCHES RELATING TO 



further criticizes Lankester's descriptions and figures, and describes 

 what may be seen in preparations specially made to solve this pro- 

 blem. He appends copies from nature of four selected transverse 

 sections, with a schematic figure combining in longitudinal section 

 the views thus obtained. We cite his own concluding summary : — 



" In the stage preceding invagination the entoderm consists of ten 

 cells, four small and six large. All these cells divide, and the ento- 

 derm invaginates in a sagittal direction. The derivatives of the six 

 large cells, and perhaps also of some of the four small, incept albumen 

 and become albumen-cells. A number of the derivatives of the small 

 cells do not, however, undergo this metamorphosis, but retain their 

 abundant granules. Since, now, the entodcrmal tract is not limited 

 posteriorly by elements which have become albumen-cells, the small 

 richly granular cells must form after complete invagination a plate or 

 string, not only bounding the cavity of the gut behind but also reach- 

 ing as far as the skin [ectoderm]. That the string in question is on 

 either side bounded by albumen-cells and at first possesses no lumen 

 — further supports this interpretation. 



I am still, therefore, of my former opinion that one ought not to 

 speak of a ' pedicle of invagination.' Yet a string or plate exists, 

 which gives rise not only to the wall of the rectum, but also to the 

 cylinder-cells of the middle-gut." 



2. The precise mode in which the middle layer makes its ap- 

 pearance is the subject of the second topic discussed in Eabl's short 

 but important paper. As soon as the stage of twenty-four cells is 

 reached, one of the four large vegetative cells, behind the transverse 

 furrow now becoming shorter, gives rise to a small daughter-cell 

 which comes to lie in or upon the furrow, and is with difficulty visible. 

 There are now, therefore, twenty-five cells, and the large residual 

 vegetative-cell just mentioned is the first cell of the mesoderm. 



Between this cell and the two contiguous endodermal cells a pair 

 of small cells (one on each side) appears. These are formed by the 

 intrusive elongation and fission of two cells from the adjacent ectoderm. 



Each of the three large endodermal cells, companion-cells of the 

 mother-cell of the mesoderm, now becomes pear-shaped. The nar- 

 rower end, pushed towards the vegetal pole, is segmented oif like a 

 bud. At the animal pole, likewise, further segmentation sets in, 

 inaugurated by the division of the four ectodermic cells, which are 

 richer in granules than their fellows. But at this stage the vegetal 

 pole is the seat of more significant changes. 



The entire number of vegetal cells has, therefore, increased to 

 eight, or seven endodermal and one mesodermal. Next, the latter 

 divides into a pair of cells. Division of the three larger endodermal 

 cells follows. Then the two mesodermal cells are thrust back into the 

 cleavage cavity. 



Finally, Rabl compares this young of Planorhis having twelve 

 vegetal cells (ten endodermal and two mesodermal) with the embryo 

 of the rabbit, as represented by E. van Beneden. Multiply by eight 

 the cells of the former ; it woiild then have ninety-six vegetal cells, 

 no longer disposed in one but in two or three layers, just as in the 



