12 BULLETIN OP THE BUREAU OF FISHERIES. 



radial processes of chromatophores could be observed. Rythmical movements con- 

 sisting of alternating short advances and retreats of the pigment granules lying nearest 

 the peripheral ends of the radial processes of melanophores were observed in fish resting 

 quietly on a background to which they were thoroughly adapted. Again, not infre- 

 quently movements of pigment granules resembling Brownian movements were observed 

 at the peripheral ends of the pigment masses extending into the radial processes of 

 moderately expanded melanophores in resting fish. Such movements are characterized 

 by Ballowitz (1913) as "a remarkable dance." 



Changes in the distribution of the pigment granules in the xanthophores take place 

 in essentially the same manner as in the melanophores. In these chromatophores 

 pigment granules could be observed advancing toward the periphery and again retreating, 

 but not along well-defined radial lines. In general the movements of the xanthine 

 granules are more sluggish and less regular than those of the melanin granules. 



The above observations on the behavior of the chromatophores in living specimens 

 of Paralichthys alhiguttus as they undergo changes in shade or color lead us to conclude 

 that changes in the distribution of the pigment granules in the chromatophores do not 

 involve essential form changes in these bodies, but that the pigment granules advance 

 peripherally and in turn retreat toward the central area while the form of the chromato- 

 phore remains essentially unchanged. 



Further evidence which indicates that the radial processes of the chromatophores are 

 not extended and retracted with each corresponding change in the distribution of pig- 

 ment is afforded by the fact that not infrequently as the pigment becomes concentrated 

 in the central area of the chromatophore aggregates of pigment granules remain far 

 out in the radial processes. If the radial processes were actually retracted, all the 

 pigment granules would necessarily be carried toward the center with them. It might 

 be suggested that these aggregates of pigment granules which remain at or near the 

 periphery break through the surface film of the cytoplasm as the radial processes are 

 withdrawn and are again engulfed as the latter are extended. However, this seems 

 highly improbable. 



The observations on the changes in the distribution of the pigment granules in 

 chromatophores recorded in the preceding paragraphs substantiate similar observations 

 made by Ballowitz on chromatophores in various species of teleosts. Ballowitz (1913) 

 has, furthermore, described very delicate radial canaliculi in the chromatophores within 

 which, he maintains, the pigment granules advance and retreat. The writer must confess 

 he has not been able to observe such canaliculi in the chromatophores of Paralichthys 

 alhiguttus in either living or fixed material. Furthermore, the lack of any regular 

 arrangement of the pigment granules in the xanthophores speaks strongly against the 

 occurrence of radial canaliculi in these bodies. 



INNERVATION OF CHROMATOPHORES. 



Another important question regarding the functional activity of chromatophores 

 involves their relationship to the nervous system. The results of operations on the eyes 

 and nerves of fishes by Pouchet (1876), Secerov (1909, 1913), Sumner (1911), and 

 others indicate that the stimuli which normally call forth changes in shade or color 

 which involve widespread changes in the distribution of the pigment granules in the 

 chromatophores are received through the eyes. The experimental work of Mast (1914) 



