108 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 52 



when frightened. If the water is changed, these pale fish soon re- 

 gain their natural color and respire slowly and regularly."^ 



According to Gage, "If one be carefully watched at a considerable 

 distance, the respirations, indicated by the alternate opening and 

 closing of the mouth and gill fissures, will be seen to take place about 

 forty times per minute. Now if one suddenly moves up very near 

 the fish, not the slightest motion of the body or of the respiratory 

 apparatus can be detected. If, however, one remains perfectly still 

 for about half a minute and watches the gill-covers, he will see them 

 commence to rise and fall very gently, and in two or three minutes 

 the respirations will be as vigorous as ever. This experiment may 

 be tried over any number of times and always with the same result. 

 This is equivalent to holding the breath with the higher animals, and 

 is apparently for the same purpose, viz, to escape detection." 



One of the means of defense resorted to by the Millers-thumb is 

 the ])uffing sideways of the head and the consequent extension of the 

 preopercular spines. This may not only deter an enemy, but may 

 entail serious consequences on one that attempts to swallow the 

 little fish. Birds have been found dead with a Millers-thumb sticking 

 in the throat. 



The species are noted for voracity, and they are indiscriminate 

 feeders. They are most active in search of food during the hours 

 of darkness, as has been remarked by Fatio. "Insects, worms, gam- 

 maroids and other small crustaceans, or the fry or even the small 

 fishes of no inconsiderable size" have been noted by Smitt and 

 others as subjects of capture. They are even cannibalistic and do 

 "not object to eating smaller brothers and sisters." 



L. Lepinay (1907) records an instance of two individuals which 

 had seized on the same victim, and the smaller, refusing to let go, 

 was taken in by the larger one as far as the head, the greed resulting 

 in death to both. When two or more fishes seize the same object 

 there is a regular tussle and pulling to and fro, which reminds the 

 observer of a couple of dogs tugging at a string. 



Girard (1851) found only "insects and larvae" in the stomachs of 

 fishes he dissected. Six specimens, taken in southern Illinois and 

 examined by S. A. Forbes, had eaten only animal food, about one- 

 fourth of which consisted of fishes, one of which was furnished with 

 ctenoid scales. Undetermined aquatic larvae (thirty-six per cent) 

 and other insects, were estimated at forty-four per cent of the food. 



'The changes of color have been also especiallj' noticed by Fatio (1882, p. 

 Ti6, 117). 



