Pyrosoma Spinosum, Herdman 221 



but were thinned away and slightly contracted, the zooids being 

 continued to the extreme margin. The substance of the test is 

 gelatinous and rather firm to the touch, but easily torn. It is almost 

 colourless, but the whole colony has a reddish tinge, due mainly to the 

 colour in the stomach and intestine of the zooids. 



The outer surface of the colony is covered with short, sharp spines, 

 each spine being situated ventrally to the oral aperture of an ascidio- 

 zooid, and sloping towards it. The spines are keeled ventrally and 

 laterally, and are about 1 millimetre in height. The inner surface 

 of the colony, as mentioned by Herdman, is smooth and glistening. 



The ascidiozooids are scattered closely through the test in no 

 apparent order, the old and young being irregularly mixed together. 

 They are very readily detached from the test, a considerable number 

 falling out, through the internal aperture, every time the specimen 

 was handled. 



The general appearance of the individual ascicliozooid is best 

 explained in the accompanying figure, drawn from a full-sized 

 individual. The prebranchial space is very small, but varies with the 

 amount of contraction. The cloacal space is small in full-grown 

 individuals, but in young zooids is considerably elongate, forming 

 almost one-third of the total length. 



The oral aperture is usually very much contracted, and sometimes 

 apparently completely closed. It is fringed by a circle of tentacles, 

 about sixteen in number, irregular both in shape and size, and a single 

 long ventral tentacle rising from a cushion-like base. The ventral 

 tentacle is usually held vertically, projecting slightly from the mouth 

 opening, and is excluded when the mouth is completely closed. The 

 other tentacles are folded inwards. 



The atrial opening is not circular, as is usual in Pyrosoma, but 

 pear-shaped, rounded above, and with the sides of the opening approxi- 

 mating for some distance below before they unite. This peculiar 

 arrangement seems to be due to the absence of muscle-fibres on 

 the ventral side of the opening, which allows the atriopore to split 

 ventrally. The pear-shaped opening was, however, found in zooids 



