478 RICHARD ASSHETON. 
cell of the ectoderm of the frog’s tadpole ciliated, yet for a 
certain period ciliated cells are so numerous as to render it 
legitimate to speak of the whole surface as being ciliated, e.g. 
tadpoles 3 mm. to 10 mm. This, however, is not the case 
from the beginning. 
The first sign of ciliation is along the edges of the neural 
plate. This is followed by a ciliated patch on the antero- 
ventral surface, in the centre of which there arises later the 
stomodzeal depression. Shortly afterwards the whole surface 
of the embryo becomes ciliated, but the above-mentioned areas 
remain recognisable by reason of the greater intensity of the 
ciliation upon them. 
If for the moment we omit the consideration of the func- 
tion of these specially ciliated tracts, it is possible to draw a 
comparison between the condition just described and the dis- 
tribution of the cilia upon the free-swimming larvee of certain 
Kchinoderms. 
The ciliated edges of the neural plate (which tract is subse- 
quently to be recognised as the cause of the naso-branchial 
current of water) might be compared with the longitudinal 
ciliated band, while the antero-ventral patch of the stomodezal 
region might be compared with the adoral ciliated band of the 
Auricularia larva of Synapta digitata described by Semon 
(18). The comparison between the edge of the neural plate 
of Vertebrates and the longitudinal ciliated band of the Auri- 
cularia and Tornaria has been made by Garstang (5), and the 
fact that the edges of the neural folds are ciliated in at any 
rate one Craniate certainly supports his suggestion. 
Garstang concluded that the longitudinal band of cilia were 
“ practically homologous with the medullary folds of the 
Vertebrata.” I am not sure what he means exactly by the 
“ medullary folds.” In Rana temporaria, up to the time of 
the closure of the neural folds, it is only the edges—that is to 
say, that part of the neural plate which does not actually form 
part of the tubular nervous system—that are ciliated. The 
actual neural plate is not ciliated until later. 
If, therefore, the ciliated ridges of the frog embryo may be 
