VIVIPAROUS FISHES OF THE PACIFIC COAST. 
455 
come in contact with them they are whirled about so that the presence of cilia becomes 
very evident, in later stages they become conspicuous in sections and smaller cilia 
seem also to be formed in the wider parts of the intestine, but I have not been able 
to demonstrate them in sections. 
Cilia have recently been described in the intestine of fishes by McIntosh and 
Prince, 1891, p. 772 : “ Many preparations show a lining apparently of cilia, and there 
is thus great probability that the enteric tract — the oesophageal portion at least — of 
young teleosteans is ciliated.” 
The intestine is now a single tube from the hyobranchial slit to the anus, with four 
well-defined regions : the gill-cavity into which the gill-slit opens, the ciliated gullet, 
a narrow indifferent region (the future stomach), and the wide hind gut with high 
columnar cells. Bo far none of the glands to be derived from the enteron have 
appeared. The liver and air-bladder appear in the stages immediately succeeding the 
one described in the third region enumerated. The hypophysis and the thyroid gland 
are derived still later from the lining of the buccal cavity and the hypoblast extending 
forward from this region. 
Before describing the further development of the alimentary tract it will be nec- 
essary to go back and follow the modifications of Kupflfer’s vesicle which, at the last 
stage described, has disappeared. 
Historical on alimentary canal. — Hoffmann, 1882, p. 5, found the hypoblast to con- 
sist of a single layer of spindle-shaped cells whose lateral extent was not equal to that 
of the mesoderm. He found the tube to be formed from in front backward. In the 
head region the hypoblast has a great lateral extent, and it is here that two lateral 
infoldings occur. The second or lower layers so formed grow toward each other till 
they meet. The two layers lie close upon each other, so that a lumen is not evident 
from the first. The gills are formed by a still further outgrowth in definite regions 
and are developed from in front backward. He considers the formation in the region 
of the ear-capsule peculiar, since the hypoblast which is here folded first of all extends 
upward so that it lies close to the auditory vesicle. These outpushings move forward 
later when the gill formation has begun and finally break through to the outside and 
form a larval spiracle. It soon disappears. In the trunk the intestine is formed by 
two infoldings, as in the head, but the hypoblast is much more restricted laterally and 
the intestine consequently much narrower. A lumen does not appear at once, and 
this is formed in the hind gut first. 
Agassiz and Whitman, 1884, found the secondary entoderm one cell deep. This is 
divided into two masses by the chorda, but the two parts unite below the chorda later. 
About the time the blastopore closes the strip beneath the chorda becomes two or 
three cells deep. This thickened mass gives rise to the alimentary canal, but how the 
tube is formed they were unable to say. 
Henneguy, 1884, found in the trout ( Salmo fario) that at the time of the differen- 
tiation of the secondary entoderm it is composed of one or two layers of cells. (Later, 
p. 122, he says three or four.) During a stage with from 12 to 18 protovertebrae, when 
Kupffer’s vesicle is well formed, the entoderm begins to become infolded below the 
anterior end of the notochord to form the intestine. The entoderm at this time 
extends forward to in front of the auditory vesicle. It is raised on each side of a 
median line, below the auditory vesicle, and it is here that the first infolding becomes 
evident. The infolding at this time does not extend beyond the anterior third of the 
