554 
Second Stage of Development (Fig. 5—7). 
The following description is based on six embryos, with total 
body lengths varying from 7 to 9 mm., but of approximately the same 
stage of development. One of them has been modelled. This is a 
stage in which the condensations are precartilage, but have differen- 
tiated enough to show the general contour of their component parts. 
With the differentiation has come the establishment-of definite re- 
lations between the parts. The pars inferior of the membranous 
labyrinth has separated into sacculus and lagena, and the semi-cir- 
cular canals are present in the pars superior. The development of 
the bony labyrinth has not progressed as rapidly; it is still precarti- 
lage. A dense strand of cells sharply outlines the base of the colu- 
mella (Fig. 5). The rounded quadrate, shifting dorsalward lies close 
to the surface in a region latero-ventral to the otic capsule. In the 
lateral and inferior part of the former, the hollow destined for the 
tympanum is well advanced. The distal part of the dorsal region 
of the first visceral pouch is entirely obliterated and the tip of the 
remaining median portion lies close to the inner margin of the quadrate. 
The ventral outpushing, pointing laterally and slightly anteriorly, 
is well within the hollow of the quadrate. The reason for this peculiar 
form of pouch is quite evident when one notices the position of the _ 
shaft of the columella and the extra-columella; the former lies in the 
fork of the pouch and the latter occupies the region of the hollow 
just posterior to the pouch. Without this anterior deflection of the 
pouch, the path of the columella would be blocked. 
The condensation of the extra-columella is of considerable in- 
terest in this stage. The greater part of the cells in the mass have 
arranged themselves in concentric rows around the distal end of the 
shaft, thereby outlining the disk-like extra-columella. In a region, 
ınedian, and a little ventral, to the disk and directly below the ven- 
tral outpocketing of the first visceral pouch, there is a small conden- 
sation, with its central cells closely packed and its outer rows less 
regular (Fig. 6). This fuses with the lower median edge of the proton 
of the extra-columella (Fig. 7). 
In the former consideration of this problem, this condensation 
was called for lack of a letter term, the processus ventralis. Previous 
to BrenpEr’s work but little had been said in regard to it. In the last 
stage described by Noack, it was casually mentioned and figured but 
no discussion of its possible significance was given. Although BENDER 
