540 FRANCIS VILLY. 
lumen for more than a short part of its length, and never open- 
ing to the exterior. In tadpoles of a slightly larger size, 
(10 mm.), there is very little trace of the cleft in the form 
found in the previous stage; but between the hyoid and 
mandibular arches in the position where the cleft, at the 
stage just described, joined the csophagus, a small diverti- 
culum of the gut is to be seen, and this no doubt is the opened- 
out hyomandibular cleft. This pouch scarcely has any exist- 
ence of its own, appearing more as a sacculation of the wall of 
the anterior part of the first branchial cleft than as an in- 
dependent gill-cleft. 
It is at this time more ventral than dorsal in position. As 
the tadpole grows it becomes less and less apparent, until it 
altogether disappears, or at least can no longer be recognised 
as bearing any resemblance to acleft. The latest period at 
which it may be readily recognised occurs in tadpoles of about 
20 mm. It is not until the hyomandibular cleft has ceased to 
be recognisable that the first trace of the Eustachian tube 
appears. The Eustachian tube, therefore, is probably not 
formed from the hyomandibular cleft, but is an altogether 
new organ. Its development may be divided into two periods 
separated from each other by a stage, during which the meta- 
morphosis from the first to the last condition takes place. The 
first beginning of the tube may be recognised in large tadpoles 
which have attained approximately their full size, but before 
the hind legs are conspicuous or toes are apparent. My 
youngest tadpole in which this organ is present is 25 mm. in 
length. At this time the Eustachian tube is very poorly deve- 
loped, and consists of a solid rod of cells running forwards 
from the dorsal and anterior edge of the first branchial cleft 
under the palatopterygoid cartilage for a short distance. This 
rod is very thin and not at all conspicuous; but it may be re- 
cognised by its position in relation to the first branchial cleft. 
Unfortunately, I have cut no sections between this and a con- 
siderably later stage; but notwithstanding this, I have no 
hesitation in identifying the structure just described with the 
Eustachian tube as it appears in tadpoles just before the time 
