NEW POINTS IN THE STRUCTURE OF AMPHIOXUS. 265 
chord. Dr. Kossmann (for the knowledge of whose paper IT 
have to thank Mr. F. M. Balfour) was unable to find similar 
cells in the corresponding opposite or ventral region of that or- 
gan, though there tooacavity isobserved in transverse sections. 
I have found such cells in this part of the notochord also. 
Muscles of the median portion of the conjoined epipleura 
(ventral wall of the atrial chamber).—Stieda’s figure 8, Plate 
II, is incomplete, in that amongst other things he omits the 
longitudinal muscular bands (related functionally to the 
distension of the atrial chamber and the opening of the 
atrial pore). These narrow longitudinal bands appear in 
such a section as detached processes on the inner face of the 
transverse muscular mass. 
Relationship of the epipleur, atrium, celom and pigmented 
canals of Amphioxus to the cavities and canals of higher 
Vertebrata. ‘The remarkable suggestion of Professor Huxley 
to the effect that the peritoneal cavity of the higher Verte- 
brata might correspond genetically to the atrial chamber 
of Amphioxus, and might then be distinguished as an 
epiceel (formed ancestrally by the coalescence in mid-line 
of two lateral folds) in contradistinction to the ‘ schizoccel’ 
and the ‘enterocel’ of various Invertebrata, led me to 
frame, by way of deduction from what seems to me a most 
fascinating and pregnant generalisation, certain questions 
which have stimulated the observations recorded in outline 
above. The case appeared to me thus: if the Vertebrata 
are descended from Invertebrate ancestors which were pro- 
vided with a ccelom or space in the mesoblast (either entero- 
cel or schizocel) ; and if the Vertebrate’s peritoneum 
Gn spite of its schizoceelous ontogeny) does not correspond 
to that ancestral ccelom, but is a new thing, an epiceel, one 
should be able to show what has become of that ancestral 
celom. Has Amphioxus any trace of it? The answer to 
that question appears to me to be, that Amphioxus has a 
very extensive space corresponding to the Vermian cclom, 
especially developed in the post-atrioporal portion of the body. 
In tracing the coelom I came upon the pigmented canals, 
which suggested a comparison with the simplest condition 
of excretory canal seen in higher Vertebrates. 
In this connection I at once referred to Mr. Balfour’s re- 
searches on the development of Elasmobranchs, remembering 
the remarkable position in which, according to his observa- 
tions, the oviduct makes its first appearance in those Verte- 
brata, namely, externally to the somatopleur (parietal lamina 
of the splitting mesoblast). In consequence I submit the 
subjoined diagrams. 
