114 The American Naturalist. [Fete oe 
To follow the species observed separately : 
In Sciæna saturna (Figs. 1-7) the chromatoblasts are first noticed 
when the gastrula covers about one-third of the yolk; that is, 
they appear quite early. They are formed along the entire mar- 
gin of the embryonic ring. When first noticed they are slightly 
separated from the surrounding cells, and their outlines become 
well defined. They thus appear larger than the cells surrounding — 3 
them, which are closely packed and whose outlines are not sharply 
defined. They either move toward the outer rim of the embry- . 
onic ring or remain stationary, while the embryonic ring moves cae 
over the yolk. At any rate, they soon come to lie entirely in the — a 
segmentation cavity, (see Figs. 1—7). At this time they are quite 
regular in outline, with probably one or two angular prolonga- 
tions. Their depth is usually equal to that of the segmentation 
cavity, and much greater than the epiblast below them or the 
ectoderm above them. As soon as they have reached the seg- ao 
mentation cavity they migrate in it, most of them being intended 
for the embryo, while many remain on the yolk, and others cover 
the oil-globule. Se 
~ While the individual cells undergo amceboid changes, their 
locomotion is not necessarily caused, as some observers supposed, 
by their amoeboid changes. One cell, which was smaller than 
usual, was seen to move quite rapidly towards the oil-globule, with 
a motion not unlike that of ciliate Infusorians caught under a 
cover-glass; z.e., it moved quite rapidly, and then seemed to be 
momentarily arrested by some-invisible barrier, when it would 
again dartalong. When the cells are first freed from the embry- 
onic ring no color is seen in them ; but before long fine granules 
are observed, resembling in most respects the minute oil-globules : 
covering the yolk. Individually these are apparently colorless, 
but collectively they form yellow or black pigment. On the bs 
globule and embryo, and later over the yolk also, the cells hecam 
flattened, more densely pigmented, and at the same time gain w 
power of contracting the pigment to a dot (Fig. 15), or expanding 
_ it to the dendritic form of the cell itself (Fig. 17). ae 
I have not observed any other cells than the migratory ones if 
this species; if others exist, they were obscured by the i 
