234 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
Liver . — The formation of the liver begins about a day after hatching. It arises 
as a solid outgrowth from the dorsal wall of the euterou not far behind the limbs. 
Its condition in this stage is shown in Fig. 138, I, PI. civ (part of a transverse 
section through a larva 100 hours old). The cells of the solid outgrowth very soon 
arrange themselves so as to form the walls of tubes (Fig. 140, Z, PI. cv, larva of 
112 hours). As the liver increases in size it grows down between the ectoderm and 
the yolk sac (Fig. 141, PI. cv, 136 hours), to the posterior wall of which it hence-' 
forth clings. 
A word may here be said as to the general characteristics of the larvae in which 
the liver has become prominent. The yolk sac is now so small (Fig. 141) that it is 
limited to the anterior part of the trunk. The periblast layer still forms, as in earlier 
stages, a protoplasmic stratum clearly marked off from the yolk. The ventral surface 
of the yolk sac is closely pressed against the ectoderm, but elsewhere the ectoderm is 
separated from the rest of the embryo by a wide space ( h. si., Figs. 141 and 143). This 
space, which may be called the body sinus, is filled with a gelatinous fluid, which coag- 
ulates intoaloose, stringy mass much like (only less dense than) the jelly of a medusa 
bell. After coagulation the jelly exhibits an irregularly radial arrangement. The 
body sinus is apparent in embryos just hatching, but the fluid which fills it does not 
develop its iieculiar qualities until much later. 
The further growth of the liver is connected with the final disaiipearance of the 
yolk and periblast. The three transverse sections. Figs. 143, 144, and 145, PL ov 
(from a larva 160 hours old), show how this disappearance is brought about. In the 
posterior section (Fig. 143) the connection of the liver with the alimentary canal is 
shown. The only part of the yolk which extends this far back is a vesicle, o. g., which- 
I take to be the oil globule. The oil globule has long before this (Fig. 151, stage of 65 
hours), become intimately associated with the periblast, the protoplasm of which has 
grown entirely round it. In the two anterior sections (Figs. 144 and 145) the intimate 
connection between the liver (1) and the yolk (y) is obvious. The periblastic proto- 
plasm, which in earlier stages formed a definite peripheral layer is now diffused through 
the yolk, which, in consequence, takes the stain. The yolk does not stain very deeply, 
but the contrast between its present and former condition is sufficiently striking. 
The periblastic nuclei have also undergone a change. They are no longer flattened 
and they stain much more uniformly than in earlier stages. Some of them have even 
a single nucleolus, and between such nuclei and the nuclei of the liver cells there is 
but very little difference. They are also no longer exclusively confined to the periph- 
ery of the yolk. The outlines of the liver cells adjacent to the yolk can not be made 
out. From these facts it is very evident that the liver is absorbing the yolk and 
periblastic protoplasm. The process is probably akin to ordinary intracellular diges- 
tion, but I could not discover the existence of any pseudopodia, nor do I believe that 
any such exist. As well as I could interpret the sections the process is something as 
follows: After the protoplasm has diffused through the yolk, the adjacent liver cells 
become actually continuous with this deutoplastic protoplasm. The cells which thus 
establish a connection with the yolk form a feeding or absorbing surface, which, as it 
incorporates new material on its yolk side, as constantly splits off’ new cells from its 
liver side. In this way the mass of yolk protoplasm shown in Fig. 145 is, bit by bit,' 
entirely converted into liver cells. The precise fate of the periblastic nuclei I have 
not determined. That they, even in a slight measure, regain their early condition is 
II 
