1913] Holmes: Pigment Cells from Larvae of Amphibians 149 



Where it is possible to find various stages between well- 

 developed melanophores and xanthophores and the simple amoe- 

 boid cells of the embryonic mesoderm that go to form connective 

 tissue, the well-developed ehromatophores are not mere pigment- 

 containing connective tissue corpuscles. The connective tissue 

 corpuscles of a young larva of Diemyctylus, for instance, are 

 different in form and mode of branching from any of the true 

 pigment cells. Figure 15 shows a typical connective tissue cell 

 from the thin part of the tail of a Diemyctijlus larva. The 

 branches of these cells are tapering and acute instead of nearly 

 uniform and blunt as in the xanthophores. They are less numer- 

 ous, less attenuated, and much less branched than those of the 

 melanophores, and they do not often anastomose. The three 

 types of cell are not connected in the older larvae by interme- 

 diate forms, but all of them represent diverging lines of develop- 

 ment from a coimnon amoeboid cell. 



While the observations here recorded made it absolutely 

 certain that the black and yellow pigment cells of young Hyla 

 larvae have an amoeboid movement, it is of course possible that 

 in the adult the cell processes of the ehromatophores are more 

 fixed in outline, and that the changes in the distribution of 

 pigment ma>- be brought about largely by the fiow of granules 

 within the cell. The extent to which amoeboid movements occur 

 in the pigment cells of the larva, however, should render one 

 somewhat skeptical regarding the conunonly received interpre- 

 tation of the pigment changes in the adult animal. 



