LIVER CELLS FROG AND RABBIT 141 



care, and is in no way a schematic representation. The 

 actual protoplasmic framework again appears very pale and 

 little stained: its apparently strong coloration in thicker 

 sections depends rather on the deposition of numerous 

 intensely stained granules in the nodal points of the net- 

 work. Although these granules, as has been said, are 

 present in great abundance, it is not by any means every 

 nodal point that is provided with them. On the contrary, 

 one frequently observes fairly extensive portions of the 

 framework which are free from them. The true nodal 

 points nevertheless stand out fairly prominently. If also 

 a somewhat fibrous composition of the meshwork is to be 

 seen in places, it was at all events never very pronounced 

 in the preparations studied by me, and the thinness of the 

 sections excludes every possibility of interpreting the struc- 

 ture, as in any way the result of separate fibrillae being 

 glued together or superposed upon one another. This 

 possibility is still further set aside when we see that the 

 protoplasmic framework here presents also the same 

 peculiarities which we have met with so frequently. If the 

 limits of the cells be more closely investigated (for which 

 purpose, of course, the thinnest and best portions are to be 

 chosen), it may be plainly seen that the two most external 

 layers of meshes of cells that border upon one another 

 are placed vertically to the line of limitation. Even in 

 Photograph VIII. this comes out plainly in places. The 

 line that marks the limit between two cells is always 

 stained very darkly, which in part, at least, may depend 

 on the abundant deposition of strongly stained granules. 

 In any case, however, a pellicle-like modification of the most 

 external lamella of the meshwork is also present. As Fig. 

 1, Plate VII. shows, this line of limitation had also fre- 

 quently a distinct zigzag outline, corresponding to the con- 

 tiguous meshes of the two limiting layers ; at times they 

 appear quite interrupted in places, so that the protoplasmic 

 frameworks of neighbouring cells pass directly into one 

 another. Since, however, to trace out in detail the relations 

 that obtain at the boundaries of the cells is, for the 

 present, beyond the scope of the task which I have under- 



