parasitic within recent Madreporaria. 



251 



The cell-contents o£ the filaments are : — 



1. Grlaiiy transparent fluid ; this renders the canals often difficult of 

 distinction, and staining with carmme renders them usually visible. 



"When removed from their canals the filaments, without any other cell- 

 contents than this glairy matter, usually show a very distinct cell-wall 

 with occasional refractive granules close to it in a few places. Usually 

 there is no colour present, but in some specimens the glairy fluid is 

 tinted a faint yet bright sap-green. 



2. Dark-coloured, bro^vn or black, cytioplasm collects here and there 

 in the filaments, Avith interspaces where the clear fluid mentioned above 

 is present ; or the whole tube may be crammed with the structureless 

 and dark mass. This cytioplasm often aggregates in regular and consecu- 

 tive portions of the tube, and the intervening colourless fluid gives the 

 appearance of cell-dissepiments. 



In most tubes the dark cytioplasm, small in amount, is situated close to 

 the cell-wall, and, there being some structureless clear fluid in the axis, 

 the filaments are refractive. 



G-ranules, excessively minute, form in the coloured cytioplasm, and 

 conidia gradually develop here and there by their aggregation in all parts 

 of the filaments. Small ovoid bodies with two or three minute dark 

 molecules within them, besides a refractive fluid, are formed out of this 

 cytioplasm within the filaments, and in the enlargements, and on them, 

 and at their ends. 



No special termiual cell containing these reproductive conidia and 

 sporidia (zoospores imperfectly developed) appears to exist ; but probably 

 the rounded and elongated ends of some filaments are the analogues of 

 the terminal fructification-cells of their congeners. 



The passage of extremely minute tubes from larger filaments, through 

 whose walls they penetrate, and the presence of small ovoid bodies 

 giving out minute filaments within the parent cell-tube, are very sug- 

 gestive phenomena. They coincide remarkably with some parts of the 

 life-cycle of the SaprolegnicB^ and this resemblance is enhanced by the 

 presence in the coral parasites of the terminal filaments cut off by a cell- 

 diaphragm from the rest. Moreover the globular endings to many of 

 the filaments, or the spherical offshoots of many, greatly resemble some 

 of the parts of species of Sciprolegnice f. But whether a cell- wall cuts off 

 the globular cell in the coral parasite, I have not been able to determine ; 

 and all the evidence I have is against this being the case. 



Outside the coral are long, branching, inosculating filaments, very 

 rarely divided by partitions, and crammed here and there with zoospores, 

 which often produce filaments when still within the cell- wall. Finally 

 the oospores are large, spherical, and apparently arise from compound 

 masses or oogonia. All these details connect the parasite with the 



* Thuret, Ann. des Sci. Nat. t. xiv. pi. 22. fig. 8, 

 t Thuret, op. cit figs. 10, 11. 



