UNEVENNESS OF OCEANIC PLANKTON 535 



if they were so then the numbers of different organisms should increase and decrease 

 together as the speed of winding slackened or increased. This does not occur. Except 

 for mechanical failure the silk roll is wound on at a speed in direct relation to the rate of 

 passage of the recorder through the water. 



In addition to such considerations, however, a direct test of the patchiness of oceanic 

 plankton was made by the use of a consecutive series of tow-nets. The method, first used 

 in investigating the patchiness of the macroplankton at South Georgia, has already been 

 described by Hardy and Gunther (1935, p. 255). Two nets, exactly the same in every 

 particular, are towed alternately at the same speed and each for the same length of time. 

 As one net comes in on the port quarter the other net is lowered away from the star- 

 board quarter and vice versa ; whilst one net is fishing the other is washed down, the 

 sample collected and the net made ready for use again. At South Georgia coarse-mesh 

 nets had been used for the macroplankton; in the present series finer-mesh nets of 

 70 cm. opening were used (the N 70 nets of the South Georgia survey) ; the front part of 

 the net is of 40 meshes to the inch and the hind section of 74 meshes to the inch (see 

 Hardy and Gunther, 1935, p. 17). Each net was towed for 10 min. at a depth of 0-5 m. 

 at a speed of 2 knots. Fifty-seven consecutive hauls were made, covering a distance of 

 19 miles. The experiment was made from the R.R.S. 'William Scoresby' in the South 

 Atlantic Ocean half-way between Gough Island and Cape Town (St. WS 133) on the 

 night of June 14-15, 1927. This series comes between records 31 and 32 (see Fig. 1). 

 Fig. 18 shows the volumes of the plankton samples obtained after the salps, which are 

 shown separately, had been removed. The samples were composed almost entirely of 

 Copepoda, young Euphausiacea, Limacina, and Chaetognatha ; the comparative pro- 

 portions of these different groups were roughly estimated as percentages of the volumes 

 and filled in in the histograms. Marked fluctuations are seen not only in the total 

 volumes, but also in each of the four groups of organisms, comparable with the type of 

 fluctuations revealed by the recorder. A typical patch of salps 1 is well illustrated. 



There is then no doubt as to the patchy nature of the oceanic plankton. As pointed 

 out in the description of the recorder methods (Hardy, 1936, p. 495) the fluctuations 

 observed on the records are really less marked than those actually in the sea because 

 there is an overlap in the sampling ; each section samples also part of the water sampled 

 by the sections on each side of it. The plotted curves of fluctuations produced are thus 

 in reality smoothed curves. 



As far as the investigations have gone, fluctuations appear to be as marked in the 

 tropical waters (records 34 and 35) as in the Antarctic Zone (records 11, 12, 13, 25, 26 

 and 27). 



It may now be convenient to bring together the evidence regarding the fluctuations in 

 the different plankton groups afforded by the thirty-five records described. 



Globigerina is only represented in any numbers on records 23-26 and Radiolaria on 

 records 23 and 24. As far as the records go they suggest regions of greater abundance 

 rather widely spaced from each other. 



1 Kindly identified by Professor W. Garstang as Salpa fusiformis, Cuv. 



