246 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
represents a stage of development which is intermediate 
between the conditions shewn in figures land 3. Itis 
evident that the diagrams are not consistent with the 
statements made in the text. It is impossible to doubt 
the general accuracy of fig. 1, for it agrees with the figures 
of all other investigators, and is in accordance with the 
statements which they make. With regard to fig. 2, 
however, a comparison of it with Salensky’s own camera 
drawing of the same stage (Pl. I, stage L, uppermost 
ascidiozoid) makes it clear that the stigmata of the former 
figure are represented by the internal longitudinal bars of 
the latter ; in this diagram and in fig. 3, the mistake has 
been made of representing the stigmata as parallel with 
the endostyle when they ought to be at right angles to it. 
I have elsewhere (Proc. Roy. Soc., LI, 1892, p. 506) 
briefly alluded to this error, but an abstract given by 
Professor Herdman in the recently published “‘ Zoological 
Record for 1891” seems to indicate that the nature of the 
error has not been generally recognised and that it ought 
to be poimted out more fully. It would appear from the 
diagram in fig 2. that Salensky actually imagines that the 
endostyle alters its position quite independently of its 
anatomical relations to the other endodermic structures— 
the endostyle moving through an angle of 90° while the 
position of the stigmata remains unchanged. The well- 
known structure of the pharynx in the adult Pyrosoma, 
and the observations made by other investigators— espe- 
clally Seeliger and Lahille—upon the relations of the 
endostyle and stigmata in the early stages of development 
form the basis for the alterations which have been sug- 
gested above; and I am not aware of any evidence which 
shews that the primary relation of the axes of endostyle 
and stigmata to one another is at any stage disturbed. 
Certain adaptive changes of form take place in the buds 
