496 EDWIN S. GOODRICH. 



Examined iu this way, the " fadenzellen " are seen to con- 

 sist of a small cell-body containing a nncleus. The cell-body 

 is of somewhat irregular shape, being circular, triangular, or 

 elongated, and occasionally drawn out into a process, re- 

 minding one of the outer processes present on some Poly- 

 cheefce solenocytes (5). A neck-like region, sometimes 

 straight, sometimes curved, gradually narrows down and 

 joins the cell-body to the distal extremity of the "thread,'' 

 which is, in fact, a slender hollow tube of great length (Fig. 

 in text, p. 495, and PI. 27, figs. 1 and 4). 



The longest tubes belong, of course, to those cells which 

 are situated fui'thest from the renal canal ; they reach some- 

 times a length of 90 fx, or nearly yL- mm. The wall of the 

 tube does not appear to be as stiff as in the case of Poly- 

 chaete solenocytes ; and under the pressure of the cover- 

 glass it is often much curved. In the living animal, however, 

 I believe the tubes are always straight. The proximal end 

 pierces the wall of the excretory duct, and projects a little 

 into the lumen of the canal (figs. 3 and 4). A long flagellum, 

 attached at its base to the cell placed at the end of the tube, 

 works rapidly down the tube and far into the excretory canal 

 (figs. 3 and4).i 



Boveri (influenced perhaps by the current dogma, which 

 afiirms, in spite of all evidence to the contrary, firstW, that all 

 tubular excretory organs are of homologous nephridial nature ; 

 and secondly, that nephridia are derived from the coelom, 

 and generally, if not always, open into it) described open 

 coelomic funnels in Amphioxus, as already mentioned above. 

 Neither in the living nor in sections of preserved specimens 

 have I been able to detect any direct communication between 

 the excretory canal and the ccelom. The branches of the 

 tubule may be very numerous, of considerable length, and 

 may themselves divide, but they end blindly (fig. 1). It 

 is to these blind ends that the tubes of the solenocytes con- 

 verge, and here the wall of the canal is less loaded with 



' I was fortunately able to demonstrate the correctness of these observa- 

 tions to Prof. Boveri himself at the zoological station in Naples. 



