282 
Birge—The Crustacea of the Plankton. 
In all of the work reported in this paper and done before the 
11th of July, 1896, a single net was employed. After that 
date the net was replaced by one of silk bolting cloth, number 
16, containing about 3600 meshes to the square cm. This net 
was cut from the same pattern as the old one. In order to 
compare the two nets they were similarly mounted in the same 
frame, and a series of comparisons made to determine their 
relative coefficient. 
To my surprise the two nets showed practically the same co¬ 
efficient. The numbers caught necessarily varied considerably, 
but the average of each of two series of five pairs showed prac¬ 
tically the same number of Crustacea; the silk net catching on 
the whole about 5 per cent, less than the old net. It did not 
seem necessary therefore to alter the coefficient of the dredge 
with the change of the net. On the 20 th of August the dredge, 
with all its appurtenances, was lost by the accidental breaking 
of the line, and the work for the remainder of the year was 
done with a similar instrument of smaller size, having a square 
opening of 100 square cm. The coefficient of this net was de¬ 
termined by comparing it with the tube, one set of comparisons 
being made by determining the number of the Crustacea. A 
second set was made by determining the bulk of the plankton 
caught by the tube and net when allowed to settle for the same 
length of time in similar tubes. Two other determinations 
were made by Hensen’s last method. (Hensen, ’95, p. 92.) The 
net was fitted with a cover having an opening of 2.5 square cm. 
Ten successive hauls of the net were made with the small open¬ 
ing and their contents mingled. This was preserved and allowed 
to settle and compared with the amount of plankton caught with 
the full opening of the net, the two quantities being similarly 
preserved and allowed to settle in similar tubes. The result of 
these three methods of determination of the coefficient of the 
net was substantially identical, the coefficient varying from 1.81 
to 2.04. The coefficient 1.9 was selected, and as a result the 
catch of this net is multiplied by 190 in order to give the num¬ 
ber of Crustacea per square meter of surface area. 
An important question has been raised, first by Hensen (’87, 
p. 12) and especially by Kofoid (’97, p. 11) regarding the vari- 
