HYPOTHESIS OF ANIMAL EXCLUSION 301 



the water previously bore a rich phytoplankton which at the time of observation had 

 died down (see p. 84). We will now attempt to correlate all the zooplankton organisms 

 which occur in sufficient numbers with the distribution of phosphate values. 



All the stations in the South Georgia survey of December-January 1926-7 have been 

 arranged in descending order of phosphate values, ranging from 113 mg. per m. 3 at 

 St. WS 26 to 74 mg. per m. 3 at St. WS 49. We will deal first with the macroplankton 

 taken in the N 100 H nets. Three of these nets, it will be remembered, were towed 

 horizontally and simultaneously at each station for one mile, one net being towed just 

 below the surface at about 5 m. depth, another at a depth of about 50-75 m., and the 

 third at a depth of about 100-150 m. The two lower nets were closed at the end of their 

 mile tow before being hauled to the surface. In the present calculations the total number of 

 each organism taken in all three nets has been used. The stations have now been grouped 

 in six ranges of phosphate values: those above 105, 100-5-105, 95-5-100, 90-5-95, 

 85-5-90 and those below 85. (Actually the five stations below value 85 all have values 

 not exceeding 80, viz. 80, 80, 79, 78 and 74.) In Table LIV the average number of each 

 organism for the stations in each phosphate range is given. The highest or higher of 

 these average numbers for each organism are shown in heavy type. We see that if one 

 exceptionally high catch of Eiiphansia superba and one of Enkrohnia hamata are omitted, 

 the majority {Calamis acutiis, Rhincalanus gigas and Antarctomysis maxima, being 

 referred to presently as interesting special cases) show an optimum concentration 

 about the phosphate values 90-5-100. That is, smaller numbers are in the regions of 

 high phosphate content (and therefore low phytoplankton production) and also smaller 

 numbers in regions of low phosphate content (and therefore high phytoplankton 

 production). We have seen that a difficulty is to be contended with in making such 

 correlations, in that some organisms make such a marked diurnal vertical migration that 

 more may be taken in the nets during night stations than during the daytime. This 

 does not apply to all organisms, since some hardly sink below the level of the lower net 

 in the daytime, but it undoubtedly applies to others. This being so, these apparent 

 correlations may be fallacious, since the proportion of night stations to day stations 

 happens to be higher in the groups of stations ranging in phosphate value from 90-5 to 

 100 mg. per m. 3 than in those where the range is above and below these values. The 

 number of night stations in each range of phosphate values is given in the table. 



To see if these correlations are in this way fallacious or not we may divide the 

 stations into two series, those taken in daylight and those taken in hours of darkness. 

 Sunset and sunrise are taken as divisions between light and darkness. Because of that 

 other difficulty we have to contend with in making correlations — namely patchiness of 

 the plankton and the smaller number of stations that we can now average — we have 

 taken for comparison wider ranges of phosphate value and divided the stations into three 

 groups of as nearly equal numbers as possible. In the daylight series we have taken 

 ranges of phosphate values over 100 mg. per m. 3 , 90-100 mg. and less than 90 mg., 

 giving groups of 10, 10 and 9 stations respectively; in the night stations we have taken 

 phosphate values of over 97, 92-97, and below 92, giving groups of 6, 5 and 5 stations. 



39 



