498 EDWIN 8. GOODRICH. 



purpose. The flagellum would serve to propel the fluid into 

 the excretory canal and thence to the exterior. That a con- 

 siderable amount of fluid could pass through the tubes be- 

 comes evident when we remember that in a full-grown 

 Amphioxus there are, roughly speaking, 100 kidneys on each 

 side, or some 200 in all. Now each of these has, on an average, 

 about 500 solenocytes — to take a low estimate, — making the 

 number of solenocytes in the whole animal roughly 100,000. 

 The average length of the soleuocyte tube may be taken at 

 2*0 mm., or 50 fx. There are, therefore, about 5 metres of this 

 thin-walled tube in each full-grown individual, representing 

 no inconsiderable area for osmotic exchange in an animal of 

 such small bulk. 



Conclusion. — I do not propose in this paper to enter 

 into a detailed discussion of the homology and taxonomic 

 importance of the segmental kidneys of Amphioxus, but 

 the extraordinary resemblance they bear to the nephridia I 

 have described in the Nephthyidte, Glyceridae, and Phyllo- 

 docidge must be insisted upon. For the purpose of com- 

 parison, figures are given of the inner end of the nephridium 

 of Phyllodoce Paretti (figs. 2 and 5), which, of all the 

 Polycha3tes I have studied, most closely resembles Amphioxus 

 in the structure of its renal organ. The nephridium of this 

 beautiful Phyllodocid is large enough to be dissected out. 

 It is then seen to end in the coelom in a bunch of blind 

 branches, which are provided with a number of solenocytes 

 arranged like the ribs of a fan. The tubes are in double 

 rows, while the cell-bodies of the two rows of solenocytes are 

 closely packed together, and wedged in alternately.^ In 

 fig. 5 I have given a diagrammatic representation of the 

 extremity of a branch of the nephridium, to compare with 

 the similar diagrammatic figure of a small portion of the 

 kidney of Amphioxus (fig. 4). These figures bring out 

 clearly the wonderful likeness of the two organs. 



* I estimate the number of solenocytes in a Phyllodoce Paretti roughly 

 at 600,000, there being about 1500 to each nephridium, and some 200 

 segments. 



