Following the usual procedure at our laboratory, the volume of all organisms less 

 thcin 2 cnn. in length plus the volunne of organisms 2 to 5 cm. in length that might be considered 

 of significant nutritional value (such as the annelids, crustaceans, cephalopods and fish) were 

 combined to give a single volume measurement for each sample. This figure was divided by the 

 estimated amount of water passing through the net to give the volume of zooplankton per unit of 

 water strained. 



After the volume measurement the contents of each sample were spread out in a 

 shallow rectangular dish (10 x 15 cm. |. A Wolffhuegel counting plate was placed over the dish and 

 all organisms greater than 0.2 mm._' in their longest dimension were counted in 10 fields, each 

 1 cm. square. Organisms larger than 2 cm. had been previously removed from the sample and 

 were counted separately. The counts of 12 major groups of the zooplankton, together with a mis- 

 cellaneous category, including such things as annelids, shrimps, and fish, are given in table 3. 

 An identification to species level was made of most of the organisms appearing in the six samples 

 collected at stations 3 and 13. These results are given in table 4 and discussed in a later section 

 (p. 14). 



DIURNAL VARIATION 



The importance of diurnal variation in estimating zooplankton abundance in the 

 central Pacific has been discussed by King and Demond (1953) and King and Hida (1954). It was 

 evident from initial examination of the volunnetric data on which this report is based that the hour 

 of hauling provided an important source of variation. This is demonstrated by the ratios of the 

 average volumes of night hauls to day hauls, which were 1. 63 for the surface samples, 0. 90 for 

 the intermediate samples, and 0. 66 for the deep samples (table 5). When the dataZ' were ex- 

 amined by means of a "t" test, we found a highly significant (P < 0. 01) difference between the vol- 

 umes of day and night samples taken at the surface, a slight but nonsignificant (Pi; 0. 5) difference 

 at the intermediate level, and a considerable, and nearly significant (P- 0,08), difference at the 

 deep level. Although the difference at the deep level may not be judged statistically significant, 

 the fact that the observed volumes of the day hauls materially exceeded those of the night hauls at 

 the deep level with just the reverse being true in the surface layer (table 5), suggests that there 

 was some vertical migration taking place and that this marked day-night difference at the surface 

 was not entirely the result of a simple dodging of the net. The lack of a strong day-night difference 

 at the intermediate level suggests that this stratum received from layers below about as much 

 plankton as it lost to layers above. 



An adjustment to remove the effect of diurnal change in zooplankton volunnes has been 

 suggested by O, E. Sette, Director, Pacific Oceanic Fishery Investigations, and described by 

 King and Hida (1954). The method is based upon the similarity between the diurnal variation in 

 zooplankton abundamce and the curve of the sine function with midnight equated to the angle whose 

 sine is +1.0. 



To examine the suitability of this method for correcting the day-night variation in the 

 surface samples of cruise 16, the data were plotted as in figure 6, the abscissa being the sine 

 function of the cingle corresponding to the time of hauling. As the values have essentially a recti- 

 linear distribution, a straight line was fitted as shown. The regression of the zooplankton volumes 

 on the sine function is highly significant (b = 0. 1553, P < 0,01), Following adjustment the night- 

 day ratio is 0.98, as compared with 1.63 before adjustment, and the variance is reduced by one- 

 third (table 6). From this evidence it appeared that the sine method of adjustment as devised for 



^/ It should be noted that this differs from the lower limit of 0.5 mm. used by King and Demond 

 (1953) on plankton from, hauls made by nets of coarser (30XXX grit gauze) mesh. 



4^/ Transformed by means of logarithms since the standard deviations of the untransformed data 

 were proportional to the means. 



