THE LYMPHATIC VESSELS OF THE SEROUS MEMBRANES. 57 



relation to the development of lymphatic sinuses and capillaries ; 

 indeed, in a mesogastrium in which there is a sufficient number of 

 such endothelial vesicles in different stages of development, their 

 connection with lymphatic capillaries may be easily shown. I have 

 also shown that in a later stage they lose their cilia. Finally, I have 

 said that an endothelial vesicle may open freely on the surface, and 

 that in that case a true stoma is formed. There is no difficulty in 

 understanding this, if we remember what has been said about pseudo- 

 stomata ; it is only necessary to imagine that if, for instance, a lymph- 

 canalicular cell, one of whose processes projects freely on the surface, 

 i.e. forms a pseudo-stoma, becomes converted by vacuolation into an 

 endothelial vesicle, and the vacuole extends along the process forming 

 the pseudo-stoma, the process must finally result in the formation of a 

 true stoma. I may add here that I have observed such endothelial 

 vesicles also in male frogs and in toads. 



We shall see that the new formation of such true stomata takes 

 place to a very great extent in chronic inflammation of the various 

 serous membranes of mammals. 



According to what we know of the development of endothelial 

 vesicles, we must express the view that they always ultimately become 

 connected with lymphatic capillaries. 



