50 THE AMERICAN NATURALIST. | (VOL. XXXHL 
The account given by Goto of the segmentation of the body | 
cavity is also most unsatisfactory, and raises the suspicion of 
many stages having been missed out. At any rate, he gives no 
ground whatever for his belief that the “epigastric ” cœlom is 
only a portion of the right posterior cœlom, for he has not 
shown how he can define the latter space. The stages he 
describes recall the period in Asterina development when the 
segmentation of the cœlom, after being formed, has to some 
extent broken down; and as it is in the highest degree improb- 
able that Asterina gibbosa has a more circuitous development 
than Asterias pallida, a reéxamination of this point, based on 
more abundant material, would probably put a different com- 
plexion on the affair. 
The criticisms made above with reference to the right hydro- 
coele apply with tenfold greater force to the origin of the peri- 
hæmal cavities. These originate long before Seitaro Goto saw 
the first trace of them. As to their origin from the ccelom, not 
the slightest doubt can exist in the mind of any one who has 
seen a properly prepared section of a metamorphosing Asterina 
larva. I have figured one of these rudiments under the mag- 
nification obtained by a Leitz immersion, putting in the outline 
of every cell, in Fig. 139 of my paper ('96), and the communica- 
tien with the ccelom is no dubious slit, but a broad opening. 
This opening surprisingly soon closes up, and in later stages 
one could imagine the space to be of mesenchymatous origin, 
if one had not seen the earlier stages. 
I may say definitely that I am very sceptical as to the mesen- 
chymatous origin of any cavity. The mesenchyme is primarily 
a series of amoebocytes floating in a-cavity, which answers retro- 
spectively to the jelly of ccelenterata, and prospectively to the 
hæmocæœæle of the higher animals. That these amcebocytes 
should first coalesce, and then hollow out to form a cavity 
within a cavity, seems exceedingly improbable, especially when 
one remembers how easy it is by missing out stages to lose 
entire sight of the genetic connections of two organs. 
In conclusion, I must remark that Seitaro Goto’s work is 
7 : not suited to solve the problems he attacks. Echinoderm 
a Tarvee are most ‘difficult ober.) to handle. Bury himself (95) 
