1891.] The Genesis of the Chromatophores in Fishes. 115 
quantities of. migratory cells. I have not had an opportunity of 
reéxamining this species or Hypsopsetta since the species of Serra- 
nus were observed. 
In Hypsopsetta guttulata the color-cells appear much later and 
not nearly in such large quantities as in Sciæna. They are first 
noticed when the gastrula covers only one-half or two-thirds of 
the yolk, and the migratory ones are formed only at or near the 
union of the embryonic shield and the embryonic ring (Fig. 14). 
Numerous cells are soon after seen along the entire embryo. I 
am not certain whether they originate i» situ or whether they 
migrate to their position. Later, when the embryonic shield is 
contracted to form the embryo, these cells move toward it, and 
finally cover it. Later other cells again move out from the em- 
bryo to cover the yolk (see Figs. 15-17). 
The observations on Serranus nebulifer (Fig. 34) were not very 
complete. Only black cells are formed, and very few cells become 
free from the embryo, all of which migrate to the oil-sphere. 
In Serranus maculofasciatus (Figs. 18—28) the chromatoblasts 
were observed about sixteen hours after fertilization. There were 
at that time a few free ones on either side of the embryonic shields. 
In fifteen minutes the number of free ones on one side had 
increaséd from nine to fifteen (see Figs. 18-22). These cells 
moved rapidly away from their place of origin, and most of them 
finally, in about two hours and a half, were found on the oil- 
sphere. A few probably returned, and finally lodged in the region - 
of the head. Besides these migratory cells, there is a broad band 
- of mesoblastic cells along either side of the embryo in which color 
is soon formed. These cells never become nomadic in the seg- 
mentation cavity, but remain attached to the embryo, over which 
they are finally nearly evenly distributed. By far the greater portion 
are yellow cells, but a few being black. Before hatching these 
cells become collected into definite masses, and some time after 
: hatching they assume the remarkable condition observed in Fig. 33. 
So far as I am aware, nothing has been written concerning the 
origin of the color itself, As stated above, the color is not due 
to the color of the protoplasm of the chromatophores but to the 
aggregation of small granules, most probably oil-spherules. The 
