122 PROCEEDINGS OF THE ACADEMY OF [Feb., 



to the nephridia of D. gyrociliatus, as figured by Meyer in 1886. Each 

 nephridium is a thin-walled tube, ciliated throughout, and composed of 

 some four to six cells. Throughout its length it adheres closely to the 

 body wall, except in the case of the nephridia of the fifth pair, which 

 float more or less freely in the body cavity. The course of each 

 nephridium is winding, but its general direction is transverse to the 

 long axis of the body, the morphological inner or central end being 

 dorsal, the outer ventral. The lumen is from 1-2 u. in diameter, and 

 widest near the middle of its course. At the central (dorsal) end (fig. 

 24) of each nephridium it gradually becomes narrower, and finally 

 terminates blindly in a flat mass of granular protoplasm, containing, 

 besides one or two nuclei, gi'eenish refringent bodies, irregularly ovoid 

 in shape. Goodrich (1897) has described similar bodies in the ne- 

 phridia of Sternaspis. Toward the peripheral (ventral) extremity of the 

 nephridium, which usually lies just laterad of the ventral nerve cords, 

 the lumen again slightly decreases in calibre. The precise point where 

 the external opening is situated could in no case be determined with 

 certainty, although prolonged and careful search was made with the 

 highest powers at my command. This difficulty is caused partly by 

 the small size and lack of definiteness of the parts concerned, and also 

 by the fact that it is only possible to study the nephridia in sections, 

 since the opacity of the living animal and the close proximity of the 

 ciliated walls of the alimentary canal make a study of the nephridia 

 diu-ing life extremely difficult. It is, however, doubtless true also that 

 the external openings are usually closed, and open for discharge only 

 at intervals, as happens in the nephridia of some annelids, e.g., J^losoma. 

 Shearer (1906) finds that each of the ducts of the nephridia of D. tcenia- 

 tus terminates in a vacuole in the ventral hypodermis, but that no 

 external opening could be demonstrated. 



The nephridia forming the first pair are, as already stated, relatively 

 complex, and are also much larger, more definite in structure, and 

 more readily observed. Nevertheless, the tracing out of the entire 

 course of these nephridia is by no means a simple task, and has been 

 accomplished solely by reconstructions from sections, one of which 

 is shown in text fig. IV, representing the left anterior nephridimn, 

 as viewed from the interior of the body of the animal. The morpho- 

 logical central end (text fig. IV, figs. 25a and 26a, c.e.n.) begins as a 

 slender strand of protoplasm, attached to the lateral wall of the first 

 segment. As it passes caudad it widens out, and in the second trunk 

 segment becomes triangular in shape, thin at its edges, the external 

 edge attached to the body wall and to neighboring parts of the nephri- 



