548 W. H. LONGLEY. 



from gray to green, or vice versa, as they swam slowly from one 

 end to the other. 



It is not necessary to appeal to such methods as the last in 

 support of the thesis that the colors of tropical fishes are of an 

 adaptive character, but this was useful in determining what 

 species might be studied most profitably upon the reef, and the 

 occurrence of what types of change might be anticipated with 

 the greatest confidence. Some results of these experiments, 

 not verified in the field, are incorporated with other observations 

 in the following table, which summarizes incompletely the in- 

 formation gained regarding color change. Failure to follow up 

 suggestions springing from these preliminary tests was due in 

 part to lack of time, and in part to the relative scarcity of the 

 animals, or difficulty in observing their habits. 



The facts embodied in table 1 should be considered merely 

 representative. No other species, to be sure, has been studied 

 even as closely as those listed, but half as many more have been 

 seen making adjustments as accurately as any mentioned. Ref- 

 erence to these forms is omitted from the table, because no 

 serious attempt has yet been made to follow the gamut of changes 

 through which they run. The reserve subject to draft, whenever 

 it seems expedient to extend the list, includes the red grouper 

 (Epinephelus morio), rock fish (Mycteroperca venenosa), yellow 

 grunt (Haemulon sciurus), the ta^ngs (Teuthis caeruleus and 

 hepatus) , red goatfish (Upeneus maculatus) , sea robin (Prionotus 

 sp.4), and the razor fish (Xyricthys, sp.^). 



All the changes noted in the first two sections of the table 

 may be observed from a boat. Some of them, especially those 

 of I. bivittatus and Thalassoma, may be followed in specimens 

 that will gather about one in 3 or 4 feet of water, but the diving- 

 hood is almost invaluable on account of the ease and accuracy 

 with which it permits the investigation of the fishes' color changes 

 to be conducted. 



This piece of apparatus is essentially an inverted, weighted, 

 metal cylinder with a plate glass window in front. It rests upon 



'' One specimen only; passed from my possession l^efore being identified. 

 * One specimen only; accidentally lost before being identified. 



