PLANKTON OF THE GULF OF MAINE 79 



composition, both from season to season and from place to place; and inasmuch as an 

 understanding of the causes of the fluctuations in the numerical strength of any group 

 of marine animals would clarify the interaction of the many physical factors that 

 govern pelagic life in the sea, information along this line is never amiss. 



Quantitative data regarding the plankton run the whole gamut from the most 

 casual to the most accurate and precise, depending on the method of collection and 

 enumeration employed, which in turn depends on whether it is the absolute numbers 

 of individuals of any group that is sought or merely their abundance relatively and in 

 a rough way. Perhaps I shall not be taken to task when I add that no wholly 

 satisfactory method has yet been devised for estimating the abundance of the larger 

 and more active members of the zooplankton. 



With immobile objects such as fish eggs, or weak swimmers such as ctenophores 

 and copepods, vertical nets of the more modern patterns yield counts of reasonable 

 accuracy; but when we attempt to deal with animals whose powers of directive 

 swimming are as well developed as those of Sagittse, euphausiids, young fish, etc., 

 the certainty that some of them — it may be many or it may be few — escape the net 

 introduces an unavoidable source of error and one that is far more serious than the 

 clogging of the meshes, resulting in only partial filtration of the column of water 

 through which the nets fish, and one that must always be reckoned within quantitative 

 work. For this same reason enumerations of the plankton contained in samples of 

 sea water of known volume, collected by water bottle or by pump, a method that has 

 proved fertile for the study of the phytoplankton (p. 398), are of no value whatever for 

 any animals except the smallest. In short, any absolute census of the total plankton 

 in the open sea will, we think, long remain something of a will-o'-the-wisp. If the 

 goal be no more than a comparative (not an absolute) estimation of the amount of 

 zooplankton present in the water, these difficulties fade. 



If the same type of net is employed for all the hauls and of a mesh calculated for 

 the general size of the plankton elements for which it is intended, and if the length of 

 the column of water fished through is either known accurately or is the same on all 

 occasions, the catches will be fairly comparable one with another, and the net error 

 (that is, failure to filter perfectly) becomes secondary. If the nets are large enough in 

 diameter 31 (say half a meter or more), with filtering surfaces sufficiently extensive in 

 proportion to the mouth area, and of a shape proper for the rapid passage of water, 

 they will certainly capture a majority of the animals in their path up to the size of 

 amphipods, Sagittae, and euphausiids. In the case of the copepods, which, after all, 

 are the backbone of the zooplankton of the Gulf of Maine, the catch will be suffici- 

 ently representative of the actual population for comparative purposes, 35 even if the 

 few individuals that chance to lie near the outer rim of the mouth of the net dodge it 

 and escape. With this end in view we have, since 1914, abandoned vertical nets of the 

 Hensen pattern, with their small mouths, for a vertical net half a meter in diameter, of 

 the Michael Sars pattern; 36 and I may add that in making vertical hauls the net has 



3 * The larger the better. 



« A whole literature, from the hands of its sponsors or critics, has arisen about the reliability or the reverse of the vertical net, 

 which has been the classic engine for quantitative plankton studies ever since Hensen (1887) first sponsored it. 

 » For specifications of this pattern see Murray and Ujort , 1912. 



7.J80S— 26 6 



