244 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
composed of small flattened cells and is apt to escape observation, especially as its i 
continuity is often broken. In staj^es subsequent to that shown in Fig. 80 the outer 1 
wall is not found (Fig. 97, PI. xcix), the lens being derived exclusively from the inner ■! 
wall of the columnar cells. When the optic cup begins to form, the layer of the J 
columnar cells (c. ep.) also suffers invagination, and the concavity thus established ; i 
begins to be filled with cells {In., Fig. 97, on right side). 
In this condition the connection of the layer of columnar lens cells with the surface i 
ectoderm is maintained by a few flattened and rather irregularly shaped cells. The ' 
connection is, however, broken (left side of Fig. 97) before the lens becomes transformed _•« 
into a solid mass, which is brought about by the increasing accumulation of cells in'®! 
the above-mentioned concavity. After the invagination of the original columnar i 
layer and the subsequent cell proliferation have converted the lens into a solid spher- \ 
ical mass, the superficial cells take on a columnar shape, while the inner cells become I 
irregularly liolygonal (Fig. 108, PI. o). The further development of the lens was ^ 
not traced with any exactness, though its condition some 15 hours before hatching i 
is shown in Fig. 118, PI. cii, where the distinction between the lens epithelium {1. ' 
ep.) and the lens fibers {1. f.) is obvious. The late development has probably nothing l 
of especial interest, but the early development is a clear case of adaptation to the ' I 
great delay in the folding off of the head. 
VII. EAR; BRANCHIAL SENSE ORGAN; LATERAL LINE, 
It has been noticed in the trout that the anlage, which was supposed to develop into 
the ear, is remarkably long (Oellacher). I have found that this anlage not only 
gives rise to the ear but to a functional branchial sense organ and to the organs of 
the lateral line as well. Before the blastopore closes there is found behind the eye a 
long shallow furrow (the sensory furrow. Fig. 62, PI. xcv, s. /.) in the nervous layer 
of the ectoderm, the epidermic stratum sometimes passing over as a bridge, and some- 
times filling the concavity with a few more or less detached cells. The transverse,, 
section (Fig. 62) shows that the groove is lined by nearly columnar cells. By the timelji 
the blastopore closes, the furrow is more marked, and presents in surface view the|( 
appearance shown in Fig. 146, PI. cvi. The nervous and epidermic layers are hence- | 
forth distinctly separated, the latter always bridging over the groove. Partly owing J 
to the convexity of the dorsal surface of the embryo, and partly to the greater thick- j| 
ness of its proximal than its distal wall (see Figs. 68-71, Pis. xcvi and xcvii), the | 
sensory furrow has, when viewed from above, the aspect shown in Figs. 146 and 147, 
PL cvi; it is only the proximal wall that is obvious. At two points the furrow begins ij 
to deepen (Fig. 146, a. s., and B. s. o.), the deepening taking place downwards and in- 
wards. At these two points the auditory sac and the branchial sense organ will 
respectively be formed. Anteriorly and posteriorly the furrow with its thickened 
wall dies away in the general ectoderm. Its extent from behind the e.yes^early to 
the somites is indicated in the figure. 
In Fig. 147, PI. cvi, a further stage in the development of the sensory furrow 
is shown. The deepening of the furrow in the auditory and branchial sense organ 
regions has continued until there are now two well-marked sacs, the anterior of which 
is the branchial sense organ (d. s. o.), the posterior the auditory sac (a. s.). Between 
