266 E. RAY LANKESTER, 
The conclusions to which they point are, first, that the 
peritoneal cavity of the Vertebrate zs the same thing as the 
uth 
cm 
Fic. 2.—Schematic transverse section of Amphioxus, to illustrate the 
relations of the ccelom and atrium. N.B. The ccelom is purposely exag- 
gerated in dimension. V. Blood-vessel. Ch. Chorda dorsalis. Pgc. Pig- 
mented canal. G. Ovarium. ZC. Lymphatic space of the metapleur. 
C. Celom. Spp. Splanchnopleur. Smp. Somatopleur. A¢. Atrium. Ap. 
Epipleur. 
Fig. 3.—Schematic transverse section of embryo Elasmobranch (modi- 
fied after Balfour). ov, oviduct; mp, muscle-plate. Other letters as in 
fig. 2. 
ie. 4.—Schematic transverse section through the cephalic region of a 
younger embryo of an Elasmobranch (after Balfour). Letters as in Fig. 2. 
celom of the Worm and of Amphioxus; second, that the 
earlier vertebrate ancestors (represented in a degenerate form 
by Amphioxus) developed epipleura, which coalesced in the 
median line postorally to form an atrium ; third, that whilst 
Amphioxus retains this atrium in functional activity, the 
other Vertebrata have lost it by the coalescence of its outer 
and inner bounding-walls, respectively epipleur and somato- 
pleur. Fourth, that whilst the indications of the earlier 
historical steps of this process are suppressed in all Craniate 
Vertebrata at present investigated, yet the Hlasmobranchs do 
continue to present to us an ontogenetic phase in which the 
somatopleur and the epipleur are widely separate ; thus en- 
closing between them an epicel (the atrium in Amphioxus). 
