126 THE AMERICAN NATURALIST. [Vor. XXXIV. 
Willey's attempt to homologize a portion of one or both of 
the proboscis canals! and their pores with the ascidian neural 
gland seems even a greater exercise of the imaginative faculty, 
for these structures in Ptychodera are not only mesodermal, 
but are a definite portion of the coelom. If they be homolo- 
gous with one or both of the collar pores and collar canals, as 
seems not improbable, then they must be regarded as nephridial 
in their relationships, for Masterman has clearly demonstrated 
the nephridial nature of the collar canals in Cephalodiscus. 
Willey, however, does not claim that the proboscis canals and 
collar canals of Ptychodera have been derived from the ascid- 
ian neural gland, but rather that the gland in the ascidians has 
been modified from a more primitive condition similar to that 
in certain adult specimens of Ptychodera carnosa. How such 
a proposition can be reconciled with the facts of the develop- 
ment of the gland in the ascidians, where it is formed wholly 
at the expense of the larval neural canal, it is very difficult to 
see. Willey's statement that the * hypophysial canal" of the 
ascidian larva “opens into the medullary tube” is hardly accu- 
rate. It is morphologically a part of the medullary tube. Dr. 
Willey must have had these facts clearly in mind, for he has 
himself published one of the most valuable descriptions of the 
ontogenetic development of the ascidian neural gland.” 
The whole question of the homologies of the neural gland of 
tunicates is an exceedingly complicated and difficult one. It is by 
no means certain even that this gland is represented in the verte- 
brates by either of the two portions of that compound structure 
which is called the hypophysis cerebri, yet this homology pro- 
posed by Julin is comparatively simple, since the ascidian 
gland and one part of the vertebrate hypophysis are of neural 
origin. Willey's theory is much more startling, for it claims a 
genetic relationship between a portion of the central nervous 
system in ascidians (the neural gland) and a portion of the 
coelom (the proboscis canals) in Enteropneusta. Yet the only 
foundation for the theory which I am able to find in his paper 
1 Willey regards the condition with two proboscis pores as more primitive than 
that with one. 
2 Several further objections to the proposed homology might be urged, but it 
seems hardly wise to treat the theory too seriously. 
