Notes on the morphology of the Tunic ata. 529 



Perophora lias vesicular gland cells with a coagulum in the center of 

 the vacuole. 



Now it is hardly conceivable that each of these différent histological 

 appearances indicates a cell of a peculiar kind or function. I rather 

 believe that the differences are due to slight differences in preservation, 

 or to the cells being in different stages of secretory activity at the 

 time of their fixation. 



This much is true: the glands (in the simple Ascidians, except 

 the Clavelinidae) or portions of glands (anterior chamber in Botryllidae) 

 which show no vesicular cells with paranuclear bodies, probably differ 

 either in the character of their secretion, or at least in the details of 

 the manner of its formation , from those glands (in the second group 

 of compound Ascidians and some of the Clavelinidae) or parts of 

 glands (posterior part in Polycyclus) which do show vesicular cells, or 

 cells with paranuclear bodies. 



In Polycyclus the cells of the posterior part of the gland, which 

 show paranuclear bodies, grade off into the cells of the ganglion, for 

 it is in this region that the gland and ganglion are fused. This sug- 

 gests that the paranuclear bodies may be related to the "centrosomes" 

 and surrounding "archoplasm" which Hunter i) has described in the 

 ganglion cells of Styela aggregata {"Cynthia partita"). With this 

 thought in mind, I examined ganglion cells of the ventral cord of 

 earthworms (Ällolohophora foetida) which had been treated in the same 

 way, i. e. killed in Perenyi's fluid and stained in Delafield's haemato- 

 xylin. It is, of course, well known that paranuclear bodies have been 

 described in the ganglion cells of the nerve cords of Annelids, and 

 have been called centrosomes. I found cells presenting almost exactly 

 the same appearance. In fact Fig. 44, Plate 37, which is a drawing 

 of one of the cells of Polycyclus with its paranuclear body, could pass 

 for a drawing of a ganglion cell from the ventral cord of the earth- 

 worm. 



Does this destroy my interpretation of these paranuclear bodies 

 as probably connected with secretion? Not necessarily, for the centro- 

 somic nature of the paranuclear bodies found in the ganglion cells of 

 so many animals [frog spinal ganglia ^), ganglion cells of Molgula ^), of 

 Ällolohophora, of Nereis^) etc.] is doubtful (cf Rohde, 1898), and even 

 if the bodies in the ganglion cells are true centrosomes the structures 

 of similar appearance in the gland cells may be connected with secretion. 



However great the differences in histological condition between 

 the two parts of the gland in Folycyclus, or between the glands as 

 a whole in différent species, e. g. in Cynthia and in Amaroecium, in 



1) Hunter, 18981. 



2) RoHDE, 1898. 



3) Hunter, 18981. 



4) Hamaker, 1898. 



35* 



