Structure <and Development of the Nephridia. 389 



cells. A cross sectiou of a similar iieplividium is represented in 

 Fig. 2G, Piate 23. The cells are larg-e and few in uumber, aud 

 possess larg-e nucleolated nuclei. At the narrower posterior region 

 of many nephridia of this stage the lumen may remain intracellular, 

 even although its anterior region may he intereellular and bounded 

 by two or three distinct cells in cross section. As the cells iucrease 

 in number, the entire lumen becomes intereellular. The distinction 

 between the two forms of lumen seems thus purely artificial and a 

 continnous giadatiou can be traced. 



The subdivision into cells takes place apparently in simple rela- 

 tion to the increased calibre of the lumen of the organ. Accordingly 

 it is found, as growth proceeds, that the lumen becomes bordered 

 by a well-defined layer of cubical excretory cells whose number in 

 cross section is in general directly proportional to the raeasure of 

 the circumference of the lumen. Figs. 27 and 28, Piate 23 show 

 successive stages in the formation of this epithelium. Fig. 28 

 represents a cross section through the body of a nephridium of almost 

 the same stage as that of Fig. 42, Piate 25. Five cells are already 

 visible in cross section, and their inner ends are beginning to pro- 

 ject into the lumen in the manner characteristic of latter stages. 

 Figs. 45, 46 — 54, Piate 25 and Fig. 6a, Piate 22 show stili more 

 advanced nephridia with a greatly increased number of cells in cross 

 section. 



Since the length of the nephridium in Arenicola is limited, never 

 exceeding that of the somite to which it belongs, the only way in 

 which it is possible to secure the necessary increase in the extent 

 of the excretory surface is by an increase in the calibre of the 

 organ. Öince each nucleus is in direct physiological relation to only 

 a limited portion of the excretory protoplasm, the iucrease in calibre 

 entails a corresponding increase in the number of cells in cross 

 section. The cells retain an approximately uniform size (as may 

 seen by comparing the above figures), and increase in number 

 (apparently by direct di vision) as the epithelium increases in extent. 

 Mitoses are never seen in the excretory epithelium; and it seems 

 clear that the cells when they exceed a certain limit of size, uudergo 

 a simple subdivision, preceded by an amitotic division of the nucleus. 

 I bave not directly observed the several stages of tbis process; the 

 nuclei, however, are frequently seen in an irregulär or constricted 

 form, and subdivision probably simply represents the extreme of 

 such a condition. The formation of a celi wall between the two 



