METHODS AND DATA. 7 



inseparable from work of this kind at sea. However, with reasonable care such cases will be few in 

 number, and will not materially affect the question of large pei'centages. 



In some groups (Chretognatha, Ostracoda, &c.) an attempt will be made to give a quantitative 

 estimate of the specimens captured, with a view to the study of their vertical oscillation. It is not 

 contended that every single specimen of a minute Ostracod or DoUolum has been picked out of the more 

 abundant hauls ; the clots which Phjeodarian Uadiolarians often make, which cannot be disentangled 

 without wrecking the specimens, are alone enougli to prevent this — still, it has been attempted. Tlie 

 population-figures presented will therefore not be absolutehj accurate, but will be, I believe, fairly 

 proportionately correct. In any case, the error is greatest in the more populous hauls, and falls to 

 zero in the less populous ; its result will therefore be to underestimate the difference between abundance 

 and scarcity of a species — if the figures were absolutely correct, the argument from the numbers 

 captured would be stronger. 



On pages 53, 54 will be found examples of a " time-depth " table, issued to the natu- 

 ralists who are kind enough to report upon individual groups. It was drawn primarily 

 with the idea of seeing whether a species rises by night and sinks by day, or not; but it 

 serves also, as in these first instances, to exhibit graphically the horizons of capture in 

 the Epiplankton and upper Mesoplankton. The ordhiates are the twenty-four hours, the 

 abscissae are depths down to 400 fathoms (not proportional). Every " comparable " haul is 

 set down at or near its intersection point, and the hauls in which a species was captured 

 are printed in heavy type. All these hauls were made with open tow-nets, and ended at 

 the surface, except those over which an arrow is placed : these latter were made with 

 the closing Mesoplankton net, the length of the arrow indicating the section through 

 which the net was open. Thus 32 I was a haul with the Mesoplanktoa net open 

 between 400 and 300 fathoms, but 36 I began at 350 fathoms and was open to tlie 

 surface. 



It will be obvious from this table that the Epiplankton hauls were deliberately grouped 

 round noon and midnight, round the surface and the 100-fathom line, as being sujiposed 

 to be " critical." There is no reason to believe that the " diurnal " conditions, which have 

 been alleged to produce a vertical oscillation of the Plankton, differ appreciably betw^een 

 (say) 6.0 a.m. when the sun is well up, and 0.0 p.m. when he has lost his power; 

 and in the limited time at our disposal it seemed more judicious to concentrate the 

 hauls at supposed critical points than to get a complete series of single hauls at 

 every depth for every hour of day and night, which would probably not have been 

 so significant. 



I must not close this introduction without expressing my indebted uess for much practical help in 

 matters of gear to Mr. W. P. Bullivant (Messrs. Bullivant and Co.), and to Mr. M. H. Gray (Telegraph 

 Works Company, Silvertown). 



