528 OHESSWELL SllEAREK. 



attached, and from the fact that they are never seen to enter 

 the canal itself, being soon swept free again into the general 

 space of the cavity, demonstrates that the end of the canal 

 is not open. And this is also borne out by the appearance of 

 the end of the canal itself^ which shows no trace of any such 

 opening. A drawing is shown in fig. 1 of the first nephridium 

 under compression, in which the soleuocytes are seen divided 

 into these two groups as just mentioned. 



Under normal conditions the organ is not so spread out as 

 shown ill this figure. The head of the nephridium projects 

 into the body-cavity space so as to look outwards and back- 

 wards, lying remarkably close under the epidermis. If for 

 convenience of description we follow Harmer (8) in dividing 

 the nephridium into three portions ; then the solenocyte 

 body being the first part, the large thick-walled part of the 

 canal marked nep. c. 2 in the figure will form the second part, 

 while the fine delicate duct into which tliis rapidly passes 

 composes the third part (nep. c. 3). 



The second portion of the first nephridium is shorter and 

 less developed than the corresponding section in other 

 uephridia. It reaches its greatest development in the third 

 nephridium. Its walls quickly widen out, and are not so 

 granular or marked with orange pigment as in the case' of 

 the following organs. These granules are so placed as often 

 to give it the appearance of being con)posed of numerous 

 cells : of being an inter-cellular instead of an intra-cellular 

 canal. This condition is again more marked in the case of 

 the third nephridium. The third portion of the nephridium 

 is remarkable for its uniform diameter, which remains the 

 same till it almost reaches its point of termination, where it 

 then narrows down to a very fine duct. It forms by far the 

 longest portion of the nephridium, running backwards and 

 ventralwards close under the skin, to terminate in the base- 

 ment membrane of the epidermis close to a conspicuous 

 vacuole just over the line that marks the anterior border of 

 the following segment (fig. 9). In sections of fixed material 

 I have been unable to trace the course of the nephridial 



