ON THE DEVELOPMENT OF THE POND-SNAIL. 385 



this middle group of cells — the histologically changed, but in 

 coarser features unchanged, bilobed group which formed the 

 gastula's stomach. I have failed to penetrate to tlie centre 

 of this mass of cells in earlier phases, and can, therefore, not 

 explain how the structure to be described comes about. 

 What can be observed is this, that as soon as the pharynx 

 and its appendix, the odontophore^s sac, becomes well 

 marked, and the tubular structure with epithelial lining in 

 the pedicle of invagination is clearly visible, then a little 

 compression and manipulation renders clear the continuation 

 of a tube-like structure with walls formed of small cells from 

 pharynx to intestine, traversing the mass of large pellucid 

 cells (PI. XVII, fig. 21). This tubular structure is un- 

 doubtedly to be regarded as the so-called stomach of the adult 

 LymncRus. The metamorphosed gastrula-endoderm-cells now 

 lie on each side of it as a pair of grape-like bunches, and 

 long after it has becon)e well-defined these two agglomerations 

 of pellucid spheres, with their enclosing network and meso- 

 blastic coat (the tunic of fusiform cells), remain. They are 

 apparently eventually absorbed as nutritive matter by diver- 

 ticula of the alimentary canal, which give rise to the liver, 

 they themselves not giving rise subsequently to any per- 

 manent tissue. Now, it is a most important question whether 

 the cell-elements which build up the so-called ''stomach" 

 (the middle piece of the alimentary canal) arise in any way 

 from the large gastrula-endoderm-cells, or from the pharyn- 

 geal in-pushing, or from the intestinal pedicle of invagination. 

 If from this last, they would just as much, as if they arose 

 from the material of the central mass of gastrula-endoderm, 

 be traceable to the invaginated cells of the gastrula-phase. 

 On the whole, it seems probable that this is their origin ; 

 but the matter is still obscure. The analogy of other 

 MoUusca does not take us very far towards a clearing up, for 

 in all cases that I have studied tiie exact mode of origin of 

 the middle portion of the alimentary canal is equally obscure. 

 It is, however, interesting and of considerable importance for 

 a true understanding of the matter, to note that in this case 

 of LymncRus we have a large proportion of the material 

 which at one time formed the wall of the gastrula-stomach 

 left outside the permanent alimentary canal and absorbed as a 

 kind of food-yelk. The case of Pleurobranchidkmi, of which 

 an embryo is represented in the woodcut, is, in a measure, 

 parallel to this, for the two large nucleated spheres (ry) are that 

 portion of the original cleavage- product of the ^^^^ which are 

 overgrown or invaginated by epiboly. Hence they represent 

 the gastrula-stomach, and, as in Lymnaus, a middle intes- 



