twilight hauls were few in number and were omitted from the comparisono 

 The resultf show (figo 3) that for both cruises the greatest average 

 volume of zooplankton occurred at the 50-meter level in both -the day and 

 the night hauls o It is obvious that the increase in the night over the 

 day hauls, which is shown at all sample levels ^ could not have occurred 

 by a shifting upward of the zooplankton population from the 100-to the 

 50-meter levels the 150-to the 100-iaeter levels etCo It would appear 

 that the greater volumes of the night hauls at all levels above 300 

 meters may be explained by the migration of zooplankton from below this 

 depths ^y "the plankton's escaping the net to a much greater extent dur- 

 ing daylight hours , or by a combination of both factors o The possibil- 

 ity that this difference in catch rate is not the result of an elaborate 

 diurnal migration^3^ but rather of a simple dodging of the net during 

 daylight hours j, as suggested by Franz (1913) » has been the subject of 

 considerable speculation on the part of plankton biologists but has 

 actually received little experimental efforto 



On cruise 9 of the John Ro Manning a sampling experiment w;as 

 conducted to determine (1) the variation between a series of day hauls 

 and night hauls at the same locality j, and (2) variations between tvro 

 localities not widely separated in di&tance and timeo The results,, 

 graphically portrayed in figure 4^ when examined by an analysis of 

 variance, indicate no significant difference (P > Oo05) between stations^ 

 no significant difference (P >0o05) between times (day or night) ^ but 

 show a highly significant (P <0„0l) interaction (table l)o This latter 

 feature results from, the fact that the day-night variation was markedly 

 different for the two localities s th© night/day ratio was lo04 for 

 stations 1 and 2j, and lo50 for stations S and 4o We cannot account for 

 this difference i weather and sea conditions varied very little during the 

 1-^day period in which the hauls were made and without a detailed count 

 of organisms in the samples ^ there were no apparent differences in co3i= 

 position. The close correspondence within each series of four hauls is 

 further assurances, hov/everj, that the method of hauling is capable of pro- 

 ducing repeatable results o 



Another experiment conducted on cruise 9 of th© John Ro 

 Manning was designed to measure differences among oblique hauls to three 

 depths J 200;, 300 j, and 500 meters j, with samples taken at approximately 

 hourly intervals over a 24-hour per led » The hauls were made by rumiing 

 the vessel between drifting buoys which were lighted at nighti therefore, 

 discotmting v/ind, we were sampling the same surface water mass through- 

 out the 24-hour period, although there ^vas a westerly drift of about 30 

 miles during this timeo The results sho'jra in figure 5, when tested with 

 an analysis of variance (table 2), yielded the following conclusions g 

 (1) differences among times of hauling^ with the 24-hour series divided 

 into four 6-hour periods, were highly significant (P < 0o0l)s (2) differ- 

 ences among depths were also highly significant (P <0„01)| while the 

 interaction or sampling error was non-significant (P >0o05)o The latter 

 indicates that the variation with time followed the same pattern for all 



3/ The phenomenon of vertical migration has been comprehend ive!ly re- 

 ~ viewed by Kikuchi (1950) and more recently by Gushing (1951) „ 



