PELAGIC PLANT LIFE 359 



practise his net could not filter the whole of the water which 

 ought to pass through ; it was possible, however, to work out a 

 coefficient for each size of net, namely a fraction indicating 

 what proportion of the total quantity of water had actually been 

 filtered. Hensen trusted chiefly to vertical hauls, since he was 

 anxious to know definitely the exact distance through which 

 the net had passed. He lowered his apparatus open, but with 

 a heavy weight attached, so that it went down end-first and 

 therefore caught nothing until hauling in began. Initial investi- 

 gations aimed at ascertaining the total quantity of plankton in 

 the photic zone, and accordingly the net was drawn in one haul 

 from a depth of 200 metres right up to the surface, or from the 

 bottom to the surface in water shallower than 200 metres, the 

 idea being to find out the quantity of plankton in a column of 

 water of known depth i metre square. 



It is not, however, sufficient merely to compare the total 

 quantity of plankton present in different localities ; it may be 

 just as important to know what there is at different depths, not 

 only because we have to consider the effect of light, let us 

 say, upon plant production, but because there may be layers of 

 water, such as we find especially in coastal areas, totally distinct 

 in hydrographical characters, and with different conditions of 

 existence. Hensen made vertical hauls from different depths, 

 and had recourse to subtraction when estimating the plankton 

 of the deeper layers, but since that time closing-nets have been 

 introduced, and we are able now to get samples from any layer Petersen's 

 we wish to study. C. G. J oh. Petersen constructed a closing- fp'^'Jft^,, 

 apparatus to go with Hensen's vertical net, and Nansen also 

 designed a vertical closing net which was invariably used by the Nansen's 

 '' Michael Sars," and found to be handy and reliable. Provided ^i^^i^g"^^. 

 only the bag be long enough in proportion to the opening, it 

 will act successfully from a quantitative point of view, though 

 we did not employ it much for this purpose, as we had better 

 methods of our own for estimating quantity. Otto Pettersson Pettersson's 

 obtained his estimates of quantity by attaching silk nets a^faching^nets 

 to a large current-meter, which recorded the velocity of the to current- 

 current, and thus indirectly supplied approximate figures de- "^^^^^' 

 noting the amount of water filtered. A series of very interest- 

 ing determinations, from samples secured in this way, has 

 been described by Broch. Broch. 



The net-method was found unreliable as time went on. In 

 the first place, it does not fairly represent the total quantity of 

 plankton, since many of the smaller forms pass altogether, or to 



