Å 
b 
scientific libraries, After they had been prepared for the m . 
116 The American Naturalist, [February, 
protoplasm is colorless. The color-granules are not. found in the 
nucleus of the cells. They are sometimes scattered through the 
whole of the remainder of the cell, but can be withdrawn from 
the pseudopods of the adult chromatophore and collected ina 
small spot. It is to the ability on the part of the chromatophores 
to thus distribute or collect the color-granules that the larva owes 
its power to rapidly change color. 
The individual spherule of the chromatophores does not pos- 
sess any definite color. It is only, as has been stated, whena 
humber of them are aggregated that color is evident. These 
granules are either a secretion of the cell itself, or they are formed 
otherwise and appropriated by the cell. The process of the forma- 
tion of the granules in the chromatophores would, of course, be dif- 
ficult to follow if they were secreted by the cell. On examining the 
medium surrounding the migratory cells for a possible explana- — 
tion of the color-spherules, it was found that the epiblast was full 
of granules or oil-spherules, similar in size and but slightly, if any, 
different in refractive index. Such spherules were especi i 
abundant in Sciæna, in which there is also an unusual number 
of color-cells. Especially towards the closing of the blastopore, 
a large number are seen over the entire portion of the yolk not 
covered by the gastrula, and it seems as though the advancing 
embryonic ring were heaping them up at the entodermic pole of ; : 
the egg. nae 
I have frequently observed individual chromatophores while ee 
the segmentation cavity, and have seen them put forth pseudopods : 
and withdraw them independently of their locomotion ; but I bat 1 
never seen them in the act of appropriating -any of the spherules ioe 
of the epiblast. ie 
There is a difference between the spherules of the yellow a Z 
of the black cells. The granules of the black cells are sal = 
and less refringent. aia? 
When first freed from the embryonic ring the color-cells usually oe 
approach the typical cell in shape, but later they become ae 
and assume the dendritic form so characteristic in the larv®- all 
These observations were made while at a distance from © 
