214 PROTOPLASM 



oue, because here and there portions of the froth are in too 

 high a focus, and hence show the false reticular image. In 

 future any fine network in which the meshes are triangular 

 throughout must in general be regarded with suspicion, 

 and cannot be considered as the true structural image 

 without sure proof. As has been said, it follows from 

 what has been set forth above, that the meshes of this mock 

 network must always be triangular. Although I have 

 devoted much thought to the point, I at least am unable to 

 find anything to support the idea that quadrangular or 

 polygonal meshes may arise in this way. Of course the 

 circumstance must be taken into account that from various 

 reasons individual connecting filaments may appear less dis- 

 tinctly marked out, and therefore may easily be overlooked, 

 in which way meshes may be formed which are apparently 

 polygonal. 



In any case it is impossible that there could be formed 

 in the way described modifications of the network which 

 would give the appearance of longitudinal fibrils, such as 

 are so frequently shown by the protoplasmic framework, 

 especially with such distinct transitions to the usual reticular 

 framework. 



Prom the arguments adduced I consider that the possi- 

 bility of the protoplasmic framework owing its origin to an 

 optical phenomenon of this kind is quite excluded. But there 

 are still other facts that prove this. As we saw, the reticular 

 appearances described arise when the drops or granules are 

 in close contact, or when there is an extremely small 

 interval between them. Now, if the protoplasmic network 

 depended upon corresponding relations, the close apposition 

 of the granules or droplets would necessarily be evident in 

 any case with a proper focus, and therefore the image would 

 be essentially different in such a focus. This, however, is 

 never the case. Hence in my opinion the protoplasmic 

 structure cannot be produced by close apposition of 

 granules. It may, however, on the contrary, very well be 

 the consequence of a close, froth-like crowding together of 

 feebly refracting droplets, for such a structure gives, as has 

 already been shown, the appearance of a network, whether 



