ranges and percentiles for most pairs, only the 

 large increase between Nl and N2 was even mar- 

 ginally significant. The percentage of empty 

 stomachs was highest at Nl and zero at and just 

 after the peak at N4. 



Triphoturus nigrescens fed principally on 

 Pleuromamma and Euphausia spp. Intact prey 

 were recorded more frequently during N2-N4 than 

 in other periods. As in the case of N. valdiviae, the 

 apparent peak at D3 was due to a few fishes' con- 

 taining large amounts of well-digested material 

 rather than freshly taken items. 



Melamphaidae 



Melamphaes danae (Figure 2) 



Few M. danae were taken at any period. None 

 were taken at SR, and only two or three were 

 taken at SS, Dl, and D2. The data indicate a diel 

 trend similar to that of several myctophids, but 

 the//-test indicated that diel differences were only 

 marginally significant (P = ca. 0.10). If the data 

 from SS, Dl, and D2 were not included, the //-test 

 indicated significance atP = ca. 0.05 and the N4 

 and D3 values differed atP<0.05. This latter, and 

 statistically dubious, manipulation indicates that 

 the apparent trend in the data is real, but that 

 more specimens would be needed to confirm it 

 properly. 



Melamphaes danae fed on a wide variety of zoo- 

 plankton including polychaetes and chaetognaths 

 as well as crustaceans — mostly small copepods 

 and ostracods. The copepods identified were all 

 either vertical migrators or shallow-living, non- 

 migrating species. Intact items were present in 

 nighttime specimens; those from daytime con- 

 tained remains barely identifiable to general tax- 

 on. 



Gonostomatidae 

 Gonostoma atlanticum (Figure 3) 



Relatively few G. atlanticum were available 

 from four periods even though 23 additional 

 specimens from the May 1974 collections were in- 

 cluded. Still there were significant (P<0.05) dif- 

 ferences in stomach fullness over the diel cycle. 

 Median values rose steadily from SR to D3, re- 

 mained at ca. 2% between D3 and N2, and then 

 dropped sharply between N2 and N3. Though the 

 median for N4 was slightly higher than that for 



504 



FISHERY BULLETIN; VOL. 76, NO. 3 



either N3 or SR, the percentage of empty stomachs 

 was highest at N4 and SR, indicating an overall 

 trend for decrease during the late night. 



Gonostoma atlanticum fed on large 

 copepods — mostly Pleuromamma xiphias, Can- 

 dacia longimana, and aetideids and scolecithricids 

 of several genera — and small (<10-15 mm) 

 euphausiids. Intact prey were found in stomachs 

 from all periods, but were mostly from the period 

 between Dl and N2. The majority of the contents 

 from N3 and N4 were well digested. 



Gonostoma elongatum (Figure 3) 



Relatively few G. elongatum were available 

 from three periods and the size range of individu- 

 als used was extremely broad (26-150 mm). There 

 is evidence from past studies that fractions of the 

 population occasionally do not migrate (Clarke 

 1974), but catches from deep night tows taken 

 during the same sampling period did not clearly 

 indicate whether or not this occurred during this 

 study. 



The //-test indicated marginally significant dif- 

 ferences (0.05<P<0.10) in stomach fullness over 

 the diel cycle, and there was no clear trend in any 

 of the parameters. Medians for all periods except 

 SR (only 8 specimens) and D3 ( 10 specimens) were 

 <1%. The means were well above the medians for 

 several periods due to a few fishes' having very full 

 stomachs. Gonostoma elongatum fed on the same 

 copepods and euphausiids noted for G. atlanticum 

 and on ostracods, amphipods, small sergestiid 

 shrimps, and fish as well. Fresh or intact items 

 were noted most frequently in specimens from SR, 

 Dl, D3, and Nl. 



Photichthyidae 



Vinciguerria nimharia (Figure 3) 



Although previous evidence (Clarke 1974) indi- 

 cated that a fraction of the population of V. nim- 

 baria does not always migrate, catches of deep 

 night tows from the same cruise indicated that 

 only an insubstantial fraction, if any, remained at 

 depth during the shallow night sampling periods. 



Diel differences in stomach fullness were highly 

 significant (P<0.005). The median was highest at 

 SS and decreased steadily throughout the night to 

 nearly zero at N4 and SR. The values were only 

 slightly higher during the day, but increased 

 sharply between D3 and SS. A clearer picture of 

 changes between D2 and SS is given by the means 



