148 WwW. B. SCOTT. 
* The alimentary cavity (Urdarmhdhle) is formed by invagina- 
tion. In the head region this cavity becomes the lumen of ‘the 
permanent alimentary canal, but in the body there arises a new 
and much larger lumen. ‘The blastopore is enclosed by the 
medullary folds, and a neuro-enteric canal is thus formed. As 
Professor Benecke! has already discovered, the anus is a new 
formation. 
The visceral clefts arise as diverticula of the hypoblast 
of the throat towards the external skin, which is resorbed at 
these points. Ata much later period a shallow sinking of the 
epiblast is formed into which all the gill-slits open. It is plain, 
therefore, that the epiblast has no share in the formation of the 
gills, which arise as vascular processes of the walls of these 
diverticula. Hight pairs of diverticula arise, of which the first 
pair very soon disappear, and so far as I have been able to 
discover, never pierce the skin at all; and I could not find 
traces of them in larve more than a day or two old. The arch 
of the first pair bears no gills, but its presence is very im- 
portant for the proper understanding of the skeletal and other 
parts of the head, as well as for the settling of the disputed 
question of the systematic position of the Cyclostomata. 
Professor Huxley? has described a hyomandibular cleft in quite 
old larvee, but 1 have not been able to verify his observations 
on this point. 
The mid-gut is at first completely filled up with yolk-cells, 
which do not begin to be absorbed until the larva has reached 
a length of about 6 mm. In the front end more cells are ab- 
sorbed, in the hind parts very few disappear, and the epithelium 
of the alimentary canal is at first remarkably high, but 
the cells gradually become very much flattened. A deep fold in 
the wallof the mid-gut makes its appearance in larve of about 
7 mm., in which fold there is a special aggregation of mesoblast 
cells. This fold is the valve, and is very similar to the vaive 
of Chimera as well as to the first rudiment of the spiral valve 
in the Elasmobranchs. 
The hind-gut is distinguished from the mid-gut by the 
absence of the valve, and further by the circumstance that it 
loses the yolk-cells, which fill it, at a very early period, while the 
embryo is still unhatched, in adaptation to the function of the 
excretory organs, which already develop an opening into the end- 
gut and through the anus outwards. 
In general it may be said that the alimentary canal suffers a 
gradual degeneration in the course of development. The canal 
is relatively largest and most important in larve of 7—10 mm., 
1 Benecke, quoted by Kupffer, ‘ Zoolog. Anzeiger,’ No. 59. 
2 Huxley, ‘ Proc. Roy. Soc.,’ No. 157, p. 129. 
