522 Dr. A. A. Gray. On the Anatomical [June 1, 
less so, and in the mammal it is comparatively simple. Further, the writer 
ventures to think that if the structures are viewed as a whole, the process by 
which this simplification has taken place may be traced in its various stages. 
In the monitor the aqueduct of the perilymph opens out of the outer 
surface labyrinth at the point where the cochlea leaves the saccule. It 
curves first forwards then inwards and then backwards, thus encircling the 
neck of the cochlea. It then passes a little backwards and downwards and 
dilates into the large perilymph recess as described above. The anterior 
wall of the perilymph recess at its upper and outer part lies in contact with 
the posterior wall of the cochlea. In the bird there is no sign of the 
aqueduct of the perilymph surrounding the neck of the cochlea; all the 
portion of the aqueduct from its commencement to its dilatation into the 
perilymph recess appears to have become obliterated. But, as though to 
compensate for this, the adjacent walls of the cochlea and perilymph recess 
have become fused and an opening appears in this wall by which the 
perilymph fluid passes direct from the scala tympani into the perilymph 
recess. This is the condition found in the bird and it is also found in some . 
reptiles such as the alligator.* The perilymph recess remains in other 
respects similar to that of the reptile. That is to say, it has the large round 
window (which is really oval both in the bird and reptile) at its lower, outer 
and anterior portion, and on its upper and inner extremity there opens a very 
short channel which corresponds to the dehiscence which occupies the same 
place in the monitor and through which the perilymph in the recess mingles 
with the cerebro-spinal fluid. Indeed, among the birds examined by the 
writer this structure still retained its reptilian form in the case of the 
penguin. 
The confusion which has arisen as to what structure in the reptile and 
bird really represents the round window of mammals has undoubtedly 
arisen on account of the presence of the opening between the wall of the 
cochlea and the perilymph recess. Since it has hitherto been supposed that 
mammals do not possess a perilymph recess, it was naturally assumed 
that the opening from the cochlea into that cavity corresponded with the 
round window of the mammals. Now this opening occupies pretty accurately 
the position of the round window in mammals, but in reality it cannot be 
said to represent the round window of the bird and reptile at all. As stated 
above, the round window of these latter lies at the lower and anterior portion 
of the perilymph recess and looks directly into the tympanum. The round 
window of the reptile ultimately comes in mammals to occupy the position 
of the opening from the cochlea into the perilymph recess, and the process 
* Retzius, op. cit., vol. 2, p. 121. 
