434 FRANCIS B. SUMNER 



case with no. 6, upon the background of dark sand and white 

 pebbles (fig. 66). According to my notes, the pale spots of the 

 fish were probably more conspicuously pale two or three days 

 after the transfer of the latter to this bottom, than they were 

 several days later. 



In spite of these admissions, it must not be supposed that my 

 illustrations represent occasional or exceptional instances. They 

 naturally include, among others, my best results in the way of 

 color adaptation. But it must be stated emphatically that most 

 of those specimens which remained in health for a long enough 

 period exhibited these changes in a marked degree. That my most 

 striking results were obtained from comparatively few specimens 

 is owing to the fact that comparatively few lived for more than a 

 few weeks in the laboratory. 



Time required for these changes 



This time ranged from a few seconds to several days. A change 

 involving the almost complete withdrawal from view of the skin 

 pigments in a dark specimen probably required the longest 

 period. Some of the best effects which were obtained in this 

 direction, starting with dark fishes, were observed at the close of 

 about two weeks, when three out of five specimens which had been 

 put in together were found to harmonize very well with the some- 

 what discolored marbled bottom of the large tank. The other two 

 specimens were very pale, though still somewhat mottled. Since 

 I was absent during the whole of this period, I do not know how 

 soon the maximum effect appeared. 



In another experiment, two specimens were recorded as being 

 'far from white' after nine and ten days respectively on the marble 

 bottom, though they were 'noticeably paler' within an hour or 

 less after being transferred to the latter. On the other hand one 

 fish (no. 12) attained nearly or quite the extreme condition of 

 pallor after four days in a white jar, but this had been preceded by 

 about two weeks on various pale homogeneous bottoms. 



In general, it may be said that in the experiments with natural 

 bottoms and with certain of the patterns, the maximum effect was 



