432 FRANCIS B. SUMNER 



nearly black and white. Such intermediate shades as appeared 

 were chiefly of uncolored gray. The glaringly contrasted appear- 

 ance which was manifested in some cases differed very signifi- 

 cantly from the almost complete monotone which was frequently 

 assumed upon homogeneous backgrounds. Such natural back- 

 grounds as the variegated sands and gravels which were used 

 produced an intermediate effect. A more or less specific skin 

 pattern was evoked, but this did not commonly consist of areas 

 which contrasted very highly. 



Although the capacity of the fish to assume one or another pig- 

 ment pattern was obviously curtailed in large degree by a fixed 

 arrangement of the chromatophores, and doubtless by other mor- 

 phological conditions, the possibilities of reaction, within these 

 liniits,were certainlj^ striking. Perhaps the most startling example 

 of this diversity in spite of sameness is to be found by comparing 

 the effects of the 1-centimeter and the 2-millimeter squares 

 (fig. Ih, li, 4a, 46). An analysis of how these various appearances 

 were produced, from a merely descriptive point of view, would 

 be difficult and not very profitable. Enough illustrations are 

 here given to enable anyone to make the attempt, if he feels so 

 inclined. 



At this point, it would be well for me to qualify the impression 

 which the reader may have received from an examination of my 

 figures and text. It must be admitted that such changes as I have 

 described were not of universal occurrence. In the case of some 

 fishes, indeed, the capacity for adaptation was found to be very 

 sHght, though I do not think that it was completely wanting 

 in a single uninjured specimen. Again, some individuals were 

 found to undergo certain changes quite readily, while completely 

 baffled by others. For example, specimen no. 4 'balked' most per- 

 sistently when placed upon the coarse reddish sand (p. 425), and, 

 after six days on this material, remained in much the same con- 

 dition as it had been when on the banded black and white back- 

 ground (fig. 4/i). The pale spots were distinct, and the general 

 appearance was one of conspicuous contrasts, with little trace of 

 red. Now this fish had throughout been one of my 'star perfor- 

 mers,' and later showed that its capacity for change was unim- 



