i2 4 SCIENCE PROGRESS. 



condition of things which may be interpreted as an early 

 stage in the process of migration. 



It is impossible to come to a decision upon this im- 

 portant matter, for embryological observations upon the 

 point seem to be entirely wanting. 



Ballowitz's investigations deal partly with the phenomena 

 of movement exhibited by the chromatophores of fishes (22), 

 and partly with the details of their innervation (23). He 

 confirms Solger (26) in his observations that during the 

 contraction of a chromatophore only the inner pigmented 

 protoplasm of the processes is withdrawn, while the outer 

 protoplasmic envelope persists in situ in the form of a 

 delicate unpigmented strand radiating out from the central 

 pigmented mass. By the use of Golgi's method these un- 

 pigmented envelopes of the radiating processes may be 

 stained in all their ramifications, and can thus be shown to 

 remain permanently fixed, even during the maximum con- 

 centration of the pigment within the cell. 



Reference may here be made to Guitel's memoir upon 

 Lepadogastcr (27), in which this investigator has also de- 

 scribed the predetermined paths which the pigment takes 

 during expansion of the chromatophore. On Plate xxviii. 

 he figures a chromatophore in different stages of expansion. 



Ballowitz further describes for the first time the re- 

 markable nature of the innervation of the chromatophores. 

 The motor nerve-fibres arise from the nerve-plexus in the 

 skin, or frequently from the cutaneous nerves themselves 

 prior to the formation of a plexus. The same motor nerve 

 often supplies many adjacent chromatophores, and the 

 chromatophores are supplied with a very variable number 

 of motor branches, often one, but usually several or many. 

 Peripheral ganglion cells are absent. Near the periphery 

 of the chromatophore the motor fibre usually divides. One 

 branch generally goes more or less directly towards the centre 

 of the chromatophore, the other takes a more peripheral 

 and spiral course over its surface. Both branches divide 

 dichotomously into numerous twigs which terminate in 

 "end-swellings" in the fixed unpigmented envelope of the 

 chromatophore. Moreover, here and there a fibre pene- 



