564 W. H. LONGLEY 



included which are insignificant in themselves, but gain interest 

 through their relation to others 



Despite its bre\'ity a number of trustworthy inferences may be 

 drawn from the table. 



The five Haemulidae, which group includes Anisotremus, are 

 nocti^rnal, and given to schooling about coral heads or among 

 gorgonians during the day. Brachygenys chrysargyreus belongs 

 to the same family, schools regularly with its relatives at some 

 of their rendezvous, and scatters with them at night throughout 

 the open spaces. It may be taken after dark with a gill net as 

 it swims along shore, where it never appears in daylight. It 

 belongs to the bionomic group of which the grunts are notable 

 members. 



The snappers (Neomaenis spp.) studied select food which 

 differs markedly from that of grunts, but secure it at the same 

 time. The stomachs of 87.1 per cent of the 208 specimens 

 examined were full in the morning or empty late in the after- 

 noon; yet it is clearly an understatement to say that that pro- 

 portion of their feeding is done at night, for identifiable food was 

 found in 85.7 per cent of those taken in the morning and in only 

 8.7 per cent of those captured twelve hours later. This is equiv- 

 alent to saying that, if equal numbers were examined in the 

 two cases, 90.8 per cent of the full specimens would occur among 

 the former. 



Even this higher percentage fails to express the truth fully, 

 for both the amount of food and the number of organisms con- 

 tained in the fishes' stomachs are greater in the morning than in 

 the evening. The possibility that some whose stomachs con- 

 tain food in the late afternoon have eaten nothing since morning 

 must also be taken into consideration, but at present too little 

 is known of the rate of digestion in the snappers to permit one 

 to speak with assurance upon this point. The only observations 

 which bear upon the matter are these, that in the three hours 

 between 5.00 and 8.00 a.m. the proportion of empty stomachs 

 in the total catch increased from 14.3 per cent to approximately 

 32.2 per cent although in the interest of accuracy a slight cor- 

 rection is necessary, since the three species do not bear exactly 

 the same ratio to one another in the two cases. 



