STRUCTURE OF EXOliETORY ORGANS OF AMPHIOXUS. 497 



excretory granules^ and thinner than elsewhere. The nuclei 

 also, which are so numerous in the other regions of the 

 tubule^ are not preseut just in those parts where the 

 solenocytes traverse its wall (Fig. in text, p. 495). 



Briefly to summarise the observations described above, it 

 may be stated that in Amphioxus there is a series of 

 excretory tubules opening into the atrium, but not into the 

 coelom, and provided at their blind internal ends with a large 

 number of solenocytes. These tubules are situated " mor- 

 phologically" outside the ccclom, being covered with coelo- 

 mic epithelium ; the solenocytes alone push through into the 

 coelomic cavity. 



At each end of the pharynx the excretoi'y organs dwindle 

 in size, as already pointed out by Boveri. The tubule in 

 these regions becomes shorter, the branches become reduced 

 in size, or are not developed at all, and the number of soleno- 

 cytes becomes much less. This is also the case throughout 

 the pharyngeal region of small specimens. Fig. 3 repre- 

 sents a small portion of the kidney of a young Amphioxus 

 ol mm. long. Not only are the solenocytes not so crowded, 

 but the average length of the tubes is less than in full- 

 grown individuals (fig. 1). 



That the segmental kidneys of Amphioxus really fulfil an 

 excretory function has been amply demonstrated by Weiss 

 !ind Boveri; but the part played by the solenocytes them- 

 selves is less clear. Boveri, who remarked that their dis- 

 tribution over the wall of the coelom coincides with that of a 

 network of blood-vessels, concluded that the " fadenzellen " 

 were concerned in the elimination of waste products from the 

 blood : '' diese Zellen dem Chemismus der Excretion dienen." 

 He could not find that they took up colouring matters — a 

 result which agrees with my own observations on the soleno- 

 cytes of Polych^etes (5, Part II). It seems to me probable, 

 therefore; that, as I have already suggested for worms 

 (5, Part II), the solenocytes are concerned chiefly with the 

 elimination of fluid substances which can pass by osmosis 

 through the thin walls of the tube, well adapted for such a 



