4 Mr. A. Alcock on the 



close crowding of the very large nuclei of epithelial cells that 

 closely invest its surface — and shows a longitudinal light 

 band occupying a little more than its median third, flanked 

 by a dark band occupying on each side a little less than its 

 niarginal third ; and when the end of the filament is brought 

 into view the lateral dark bands are found to directly inoscu- 

 late round its lip. The dark marginal band is in fact a broad 

 capillary filled with blood-clot, disposed in a long narrow 

 loop. 



Under a higher power the surface of the filament is seen to 

 be uninterruptedly covered with polygonal epithelial cells in 

 the closest possible contact with one another. These cells 

 are remarkable in consisting of little but a large vesicular 

 nucleus lying within a thin and difficultly visible capsule of 

 cell-protoplasm ; the nuclei have a diameter varying from 

 7*5 to lOmicromillimetres, and are often polygonal by mutual 

 compression. 



A transverse section of a filament looks like a pair of pince- 

 nez (fig. 3), each lens of the pince-nez being formed of a cross- 

 section of a simple capillary tube, with a wall one cell thick, 

 enclosed in a frame formed by a single row of large nucleated 

 epithelial cells, and the bridge of the pince-nez being formed 

 of two rows of these cells with a layer of flat nuclei, continued 

 across from the capillary wall on each side, between them. 



To recapitulate and restate : a gill-filaraent is nothing more 

 than a long narrow loop of a capillary of wide bore with a 

 wall one cell thick, enclosed in a folded sheet, also only one 

 cell thick, of small epithelial cells which consist of little but 

 a great nucleus. 



As to the function of the gill-filaments : their vascularity 

 and the nature of their epithelium clearly indicate great 

 activity. They do not seem to have any attachment to the 

 uterine wall, but, on the other hand, the manner in which 

 they enfold the yolk-sac leads to the belief that they assist in 

 absorbing the nutrient yolk. And the irregular indefinite 

 nature of the channels of the stalk of the yolk-sac, which 

 channels, moreover, seem to carry only yolk-particles and not 

 blood, seems to give strong support to this view. 



4. The Uterine Trophonemata o/'Pteroplatsea micrura 

 in the Earlier Stages of Pregnancy. 



The structure of the nursing-filaments when in active 

 function for the benefit of the foetus has already been described 

 and figured by Professor Wood-Mason and myself [vide 

 * Proceedings of the Royal Society,' vol. xlix. pp. 359-367), 



