256 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
mesoderm plate as a rectangular mass placed at right angles to the long axis of the 
body, Fig. 98, The vertical planes of division separating it from the mesoderm behind 
and the somites in front are simple planes (Fig. 85), and in transverse section the 
somite is undivided (E’ig. 109, PI. c). All that is changed as the somite grows older. 
Fig. 98 shows that the somites become inclined to the long axis of the body, the 
division planes running from without and posteriorly; Fig. 85 that they become bent 
forwg,rds at about their middle ; and Fig. 110, PI. xovii, that each somite becomes 
constricted into a dorsal and ventral portion (corresponding to the dorso-lateral and 
ventro lateral trunk muscles). 
While the somite is undergoing the changes of shape just described it also passes 
through a histological metamorphosis. The mesoderm plate, before the formation of 
somites begins, is made up of irregularly polygonal cells, which at the surface approach 
the cubical sha^je and give the plate a smooth bounding surface. The posterior rem- 
nant of the mesoderm plate (Pigs. 85 and 98, un. mes.) preserves these characteristics, 
but the bounding cells become more decidedly columnar. A transverse section of the 
undivided mesoderm is shown in Fig. 90, PI. xcviii. When the somite is first con- 
stricted off, it is consequently made up of polygonal cells (Fig. 86, PI. xoviii, more 
highly magnified view of one of the iiosterior somites of Fig. 85, and Fig. 109, PI. 
0 , transverse section) and has on all sides a smooth bounding surface. 
During the development of the somite the polygonal cells elongate in the direction 
of the chief body axis, and it finally is brought about that each cell stretches from the 
anterior to the posterior surface of the somite. In Pig. 87, PL xcvill (more highly 
magnified view of one of the anterior somites of Fig. 85), the cells have undergone this 
elongation. Each of these long cells becomes transformed into a muscle fiber. 
When the somite is first formed it is, as I have said, undivided in transverse sec- 
tion and has on alt sides a smooth limiting surface. The smooth surface is lost on 
the i^roximal side of the somite (Fig. 94, PI. xcix), many of the cells coming to jut 
out in an irregular fashion. The next change to be described is the division into 
dorso-lateral and ventro lateral muscle tracts. The somite constricts in a plane about 
opposite the notochord (Fig. 110, PL oi), and the tracts thus formed {d. 1. m. and 
V. 1. m.) are made more distinct by the existence of the cells c. t. While the remaining 
cells of the somite (exclusive of possible migratory cells, in regard to which I have no 
satisfactory observations) become transformed into elongated muscle cells, with large 
conspicuous nucleoli, the cells, c. L. remain small, have inconsiiicuous nuclei and nucle- 
oli, and form, as shown in the figure, a dividing wedge between the two muscle tracts. 
I really do not know what becomes of these cells, since the somite retains through- 
out embryonic life the character shown in Fig. 110. Sections through the larval 
stages suggest, however, that they become transformed into connective tissue. At 
least in the larva the muscle tracts are separated by a few scattered cells (Fig. 136, 
PL CIV). 
The transformation of the elongated somite cells into muscle cells takes place 
after hatching, though, as is well known, the body is capable of strong muscular con- 
tractions before this time. Oscar Hertwig has described the metamorphosis which 
the homologous cells in Triton undergo (Lehrb., p. 270). In Triton distinct muscle 
fibrils appear in the cell protoplasm, first appearing in the peripheral part of the cell, 
but gradually forming in the inner portion also, until the whole cell is transformed 
into a bundle of fibrils. The development of the muscle cells in the Bass is of a 
