THE OSTEACODA. 29& 



In such a study as the present, where it is difficult to assimilate the meaning of rows of figures of 

 population at various depths, it seems easier to grasp the distribution visually by representing it as a 

 symmetrical plane spindle, rather than by the ordinary curve. This is done by erecting the base-line, 

 by laying off each observation on both sides of the line, and by joi;;ing the points thus obtained ; in 

 other words, two ordinary curves are set back to back in order to make a symmetrical figure. By this 

 method (which is by no means novel), the maximum intensity — for which I suggest the term plurimum — 

 becomes readily conspicuous. 



The observations plotted on the upper part of Plate 27 are : — (1) Average specimens per hour haul 

 for 0, 25, 50, 75, and 100 fm. ; the ordinates are set opposite to the depths at which the open nets were 

 towed horizontally. (2) Average specimens per hour haul for greater depths : the ordinates are in this 

 case opposite to the middle of the water-column through which the closing-nets were hauled vertically ; 

 thus the average for 400-300 fm. is plotted at 350 fm., the average for 750-500 fm. at 625 fra. This 

 seemed to be the fairest way of representing the facts *. Fractional averages less than 0*5 are represented, 

 by a thin line ; if more than 0"5 they are taken as I'D. The hauls with the closing mesoplankton net 

 are separated from those with the open horizontal epiplankton nets by a white space or line. All tiie 

 figures are plotted to the same scales. The averages per hour haul and per 100 fathom haul are of course 

 not directly comparable with one another, because the mass of water filtered by the vertical closing-net 

 is so much the smaller. 



In the lower part of PI. 27 similar diagrams, to the same horizontal scale, have been used to express- 

 roughly my interpretation of the observed facts, and have therefore only the value of a personal opinion. 

 These " distribution spindles " are obtained (1) by omitting those captures which I regard as representing 

 only dead or dying specimens on the grounds already given (pp. 282-3) ; (2) by a process of 

 " smoothing the curves " by drawing lines from the plurimum to the inferred upper or lower limit, 

 paying regard to intermediate averages in the epiplankton. In these diagrams, unlike the others, it 

 seemed fairest to extend the " distribution spindle " to the extreme possible points, in order to allow for 

 scattered specimens which the net might have possibly missed, owing to the thinning out of a species as 

 we recede from its plurimum in both directions vertically. Thus, if the highest admitted specimens were 

 in a 300-200 fm. haul, the figure has been carried up to 200 fm. ; if the lowest admitted specimens 

 occurred between 750-500 fm., it has been carried down to 750 fm. 



The same general methods have been adopted for the graphic representation of day and night distribution 

 in fig. L (p. 302), but on a doubled vertical scale. 



These diagrams show conclusively liow valueless are any generalisations as to the 

 vertical distribution of even a genus of Ostracoda from one place, day, or even hour ta 

 anotlier. Still less can such generalisations be made for a family or group. Each 

 species seems to be in a state of chronic movement, and must be studied separately, as 

 an entity with its own specific likes and dislikes ; this should be attempted at every 

 separate geographical position possible, if we are ever to understand the laws of 

 planktology. 



* The averages for the closing-net 150-50 fiti. (21^;) are therefore on tlie 100 fm. ordinates, as well as those 

 for the horizontal open nets at 100 fm. Fortunately, the averages only coiiieiJc in the case of proctra, and ',?ill 

 therefore not lead to contusion. 



SECOND SERIES. — ZOOLOGY, VOL. X. 48 



