COLOR CHANGES IN TELEOSTS 529 



with, the changes in the skin; but nothing is known concerning their 

 immediate causation and control. 



Mode of Control in Teleosts — The speed with which fishes can effect 

 skin changes, together with the fact that they occur all over the body at 

 once, speaks for nerve impulses; and in fact it is now generally believed 

 that in all poikilochromic teleosts the nervous system is in practically 

 complete charge as lieutenant to the eye. The work of Pouchet in 1876 

 was the first to indicate this. The cutting of nerves in turbots put out of 

 action the chromatophores of corresponding skin areas. In 1893 Ball- 

 owitz demonstrated profuse nerve endings on the melanophores. Others 

 since have been able to make out that these autonomic fibers are of two 

 kinds, affording a double, reciprocal innervation. There is a little 

 evidence that endocrine secretions — so nearly all-important in amphibian 

 dermal changes — play a very minor part in teleosts. It has been claimed, 

 though with insufficient proof, that posterior-pituitary extracts increase 

 the amounts of melanin in teleost melanophores, and that the lipophores 

 have no nerve supply at all and are entirely under pituitary control. The 

 isolated melanophores of a single scale will respond to autonomomimetic 

 and other drugs (though not to visible light) , but this does not imply as 

 much for a hormonal control in the intact fish as it may seem to do. We 

 now know that nerve fibers arouse effector end-organs by means of sec- 

 retions from their tips — the so-called 'neurohumors' ; and that these latter 

 include such substances as adrenalin and acetylcholine. Nervous and 

 hormonal control-mechanisms may thus be said to have a common 

 denominator. 



It has been found that depressants of the nervous system, such as 

 anaesthetics, produce an 'expansion' of the melanophores of Fundulus; 

 while reflex paling results from the administration of stimulant drugs. 

 The dermal changes of this much-studied fish are speeded up by increased 

 temperature, and proceed at different rates under the different osmotic 

 circumstances of fresh water versus salt; but these facts are not incom- 

 patible with the idea of nervous control. When a spinal nerve is cut, as 

 Pouchet originally showed, the melanophores in the skin supplied by the 

 nerve become expanded and remain so for many days, until the motor 

 fibers of the nerve regenerate. Some slight and sluggish activity remains 

 in the chromatophores however, even after their denervation, indicating 

 that the ebb and flow of hormonal concentrations in the blood stream 

 are not without some effect. Local interference with the circulation abol- 

 ishes this residual activity, though this may be due more to the shutting 



