188 J. DUESBERG. 



like the product of secretion (the very large networks, such as 

 that represented in Fig. u, may make exception and stain with 

 alizarin) ; but if we stain with safranin, we find that the secretion 

 product alone takes up the dye and stands out conspicuously. 

 A second reason which argues in favor of my interpretation is 

 the existence of all possible transition forms between the net- 

 work and the first stages of chondriolysis. In the cell represented 

 in Fig. 8 there are some of those vesicles .which, as we know, 

 result from the artificial fragmentation and swelling of the chon- 

 driosomes. There are present also some forked forms, which 

 are probably produced by the confluence of the swollen chondrio- 

 somes, and give an indication of how a network can be built at 

 the expense of the normal, independent chondriosomes. Again, 

 although at times the network appeared to be connected with 

 the intracellular accumulations of secretion, on closer examination 

 no such connection could be actually established. Finally, there 

 is a difference in the structure; the product of secretion usually 

 appears granular, while the substance of the network is homo- 

 geneous. 



As stated above, I consider this network as produced by the 

 transformation of the chondriosomes, not by a natural process, 

 as is well understood, but by an artificial one. Of any inter- 

 vention on the part of the chondriosomes in the formation of the 

 secretion product, an intervention which in a number of cases 

 seems to me entirely plausible, nothing could be seen here. 

 Under what form does the product of secretion of the interstitial 

 cells make its appearance? 'I have mentioned the presence of 

 granules that are preserved by a number of fixing reagents. 

 These granules are certainly secreted by the cell, but in my 

 opinion they are too scarce and inconstant to be considered as 

 the forerunners, or at least the only forerunners, of the product 

 of the cell. One must consequently admit that the bulk of the 

 secretion makes its appearance in a diffuse or non-stainable form. 

 I would suggest, without, however, being able- to substantiate that 

 opinion, that the product of secretion is represented by the sub- 

 stance which is responsible for the diffuse staining of the proto- 

 plasm, something like Mulon's osmiophile or siderophile substance. 

 But there is no indication in this case, as Mulon states it, that 



