TllIRTEEXTll AXXlAl, REPORT lOy 



Tlie color plates ])aimccl wil'.i tiic use (if the purlahle aciuariuin 

 may, therefore, he classed with the photographs made bv the 

 >anie method. They show ])hases of color and niarkinj^s ex- 

 hihted l\v specimens more or less affected by fright. The plates 

 reproduced from paintings of fresh, but rapidly fading dead sjjeci- 

 mens. may, if very (phckly done, show color phases resulting 

 from fright. Otherwise they can show only tints resulting from 

 the slow relaxation of the color cells as the dving fish loses con- 

 trol of them. 



While i;aintings and photographs of fishes made in the ways 

 described above niay present very well the colors and markings 

 knowii to j)rofessional fishermen and to anglers — that is, colors 

 exhibited l)y captured and frightened or dying specimens — the 

 most of those with which we are ac(|uainted do not portrav the 

 various species in the phases in which thev exhibit themselves in 

 life. 



All illustrations of fishes which are capable of sudden changes 

 in color should, with a view to scientific exactness, be accom- 

 panied with data respecting the phases of color shown, and the 

 methods used in producii'g them. Color plates should never be 

 made from dead fishes if living exam])les can be procured. It is 

 equally important that similar data should accom])any written 

 descriptions of colors. The keeping of tropical fishes in public 

 aquariums has now made possible some knowledge of their colors 

 under normal conditions. In written descriptions in general, the 

 familiar caption "Color in life," is inexact, since it usually means 

 merely the changing colors of a dying specimen. 



It does not appear that there is any phase of color in the twen- 

 ty-six species under observation, which can be called the per- 

 manent life color ; frequent changes take place dependent upon 

 activity, rest, play, anger, fright, temperature, food, light, or 

 other causes. There are usually two or three phases of common 

 occurrence, others being of less frequent appearance. The color 

 changes are not necessarily connected with the breeding season, 

 since they can be observed daily throughout the year. 



In the following notes no attempt has been made to write out 

 new svstematic descriptions of each species, the object in view 

 being merely to direct attention to the facts that all of the 

 fishes observed have diflferent colors and markings at different 

 times, that most of the phases have been hitherto undescribed. 

 that the changes from one phase to another are sudden, and to 

 state briefly the general character of each phase. 



