534 MAYNARD M. METCALF, 



to the crowding occasioned by the vesicle. According to this sug- 

 gestion, then, the duct of the neural gland (which is the same thing 

 as the nerve tube of the tadpole) lies in the morphological median 

 plane, and the brain lies to the left of this plane, because at the time 

 of its first origin, it was crowded to the left by the presence of the 

 sensory vesicle on the right side. In the later stages of metamor- 

 phosis the large and solid ganglion shifts back toward the median 

 plane, crowding the gland and its duct over to the right. Such an 

 interpretation is at least plausible and it is difficult to conceive of 

 any other factor to which the asymmetry may be due. Of course 

 the reason for the asymmetrical position of the sensory vesicle in the 

 tadpole is still to be sought. This may be a much more difficult 

 problem. 



The Fyrosomidae. 

 I have had but one species of Pyrosoma for study. This is, I 

 think, P. giganteum Les. 



Pyrosoma giganteu^n Les. 

 Plate 38, Figs. 52—54. 



The intersiphonal organs of Pyrosoma conform in general to the 

 type found in the second group of compound Ascidians, including all 

 the families studied except the Botryllidae. The conical funnel leads 

 by a short duct into the small gland below the ganglion. The cells 

 of the gland are vesicular and take haeraatoxylin stain but lightly, 

 resembling the gland cells of the second group of compound Ascidians. 

 The rapheal duct leads back from the gland for a short distance, soon 

 losing its lumen and then dwindling to a minute cord and dis- 

 appearing (Figs. 53 and 54 r.d). 



Above the rudimentary rapheal duct a small cord of cells pushes 

 out from the posterior end of the ganglion {r.n in Figs. 53 and 54). 

 In my rather poorly preserved material I can trace this cord only 

 about two-thirds of the way to the upper end of the dorsal raphe. 

 I believe that it stops here, not entering the raphe. There are a 

 few nerve fibres accompanying this cord of cells. I find also a nerve 

 arising from the ventral surface of the ganglion not far from the 

 posterior end and somewhat to the right of the median line. This 

 nerve runs past the upper end of the raphe, but I cannot see that 

 it gives any fibres to the raphe. Its course is parallel to the cellular 

 cord just mentioned, but the two are not in contact. My material, 



