ON THE NEPHRIDIA OF THE POLYCHiETA. 449 



faintly tinged with hseraoglobiu and of very irregular shape, 

 being generally covered with amoeboid processes. Normal 

 rounded hsematocytes are occasionally present, especially in 

 young specimens. The numerous processes on the hsemoglo- 

 binous cells give them a spiny appearance ; but, as a matter of 

 fact, the pseudopodia are not merely spine-like in shape, being 

 really thickenings in thin expansions of protoplasm. When 

 these cells are watched under the microscope the pseudopodia 

 can be seen to begin as little rounded knobs, which gradually 

 expand, spreading out in thin sheets, supported here and there 

 by ribs or thickenings. It is these which, on a casual glance, 

 have the appearance of freely outstanding processes (fig. 22). 

 The granular oval cells occur also in small numbers, but the 

 ordinary white amoeboid corpuscles appear to be very rare or 

 entirely absent in this species. On injecting a Glycera 

 siphonostoma with carmine or Indian ink, we find that the 

 granules are taken up in quantities by the hsemoglobinous cells, 

 just as in the previous species they are absorbed by the leuco- 

 cytes. The hseraatocytes are therefore both amoeboid 

 and phagocytal. This, so far as I am aware, is the first 

 instance in which a free cell containing haimoglobin has been 

 shown to ingest foreign particles. 



Gl. unicornis is, in respect to its ccelomic cells, interme- 

 diate between the two species described above. As a rule, in 

 Gl. unicornis normal round hseraatocytes occur in the coelom, 

 together with a number of amoeboid leucocytes and the usual 

 granular oval cells (fig. 6). Such specimens, when injected, 

 show that the particles of carmine or Indian ink are taken up 

 exclusively by the leucocytes. 



On the other hand, it may be frequently observed that the 

 leucocytes are very rare, and that the hsematocytes have a 

 more or less pronounced tendency to produce pseudopodia. In 

 accordance with this it is often found in injected worms that the 

 particles have been ingested by the red hsemoglobinous cells. 



I have not met with any evidence distinctly supporting the 

 view that the hseraatocytes are modified leucocytes, or vice 

 versa; yet there seems to be no doubt that the functions of 



