EVENING DISCOURSES. 7438 
obtain very different results according to the date of his visit. For ex- 
ample, on three successive weeks about the end of September he might 
find evidence for as many different far-reaching views as to the composition 
of the Plankton in that part of the Irish Sea. Consequently, hauls taken 
many miles apart and repeated only at intervals of months can scarcely give 
any sure foundation for calculations as to the population of wide sea areas. 
It seems, from our present knowledge, that uniform hydrographic conditions 
do not determine a uniform distribution of Plankton. [Some statistics of 
hauls shown. | 
These conclusions need not lead us to be discouraged as to the ultimate 
success of scientific methods in solving world-wide Plankton and fisheries 
problems, but they suggest that it might be wise to secure by detailed local 
work a firm foundation upon which to build, and to ascertain more 
accurately the representative value of our samples before we base conclusions 
upon them. 
I do not doubt that in limited, circumscribed areas of water, in the case 
of organisms that reproduce with great rapidity, the Plankton becomes more 
uniformly distributed, and a comparatively small number of samples may 
then be fairly representative of the whole. That is probably more or less 
the case with fresh-water lakes;' and I have noticed it in Port Erin Bay 
in the case of Diatoms. In spring, and again in autumn, when suitable 
weather occurs, as it did two years ago at the end of September, the Diatoms 
may increase enormously, and in such circumstances they seem to he 
very evenly spread over all parts and to pervade the water to some depth; 
but that is emphatically not the case with the Copepoda and other con- 
stituents of the Plankton, and it was not the case even with the Diatoms 
during the succeeding year. 
I have published elsewhere an observation that showed very definite 
limitation of a large shoal of crab Zoéas, so that none were present in one 
net while in another adjacent haul they multiplied several times the bulk 
of the catch and introduced a new animal in enormous numbers. [Diagrams 
shown.] Had two expeditions taken samples that evening at what might 
well be considered as the same station, but a few hundred yards apart, they 
might have arrived at very different conclusions as to the constitution of 
the Plankton in that part of the ocean. 
It is possible to obtain a great deal of interesting information in regard 
to the ‘hylokinesis’ of the sea without attempting a numerical accuracy 
which is not yet attainable. The details of measurement of catches and of 
computations of organisms become useless, and the exact figures are non- 
significant, if the hauls from which they are derived are not really com- 
parable with one another and the samples obtained are not adequately 
representative of nature. If the stations are so far apart and the dates 
are so distant that the samples represent little more than themselves, if the 
observations are liable to be affected by any incidental factor which does 
not apply to the entire area, then the results may be so erroneous as to be 
useless, or worse than useless, since they may lead to deceptive conclu- 
sions. It is obvious that we must make an intensive study of small areas 
before we draw conclusions in regard to relatively large regions, such as the 
North Sea or the Atlantic Ocean. Our Plankton methods are not yet accu- 
rate enough to permit of conclusions being drawn as to the number of any 
species in the sea. 
The factors causing the seasonal and other variations in the Plankton 
already pointed out may be grouped under three heads, as follows :— 
(1) The sequence of the stages in the normal life history of the different 
organisms. 
* See, however, C. Dwight Marsh, in T'rans-Wisconsin Acad. of Sci., vol. xiii., 
and Wisconsin Geol, and Nat, Hist, Survey, Bull. xii, 
