4 34 Birge—Vertical Distribution of Pelagic Crustacea. 
numbers, while not absolutely the same, were practically equiv¬ 
alent. An illustration taken at random from the cards may be 
cited: 
July 17, 12 midnight. 0-3 m. 
Observer , H. Diaptomus, 206.Cyclops, 34 
Observer , O. Diaptomus, 216.Cyclops, 25 
July 27, 12 midnight. 0-3 m. 
Observer , H. Diaptomus, 146...Cyclops, 65 
Observer , O. Diaptomus, 123 ...Cyclops, 79 
This system of checking each other’s work was carried on 
through the entire task, and whenever the numbers obtained dif¬ 
fered widely from each other a second counting was made, in 
order to secure greater accuracy. 
The results thus obtained were multiplied by a factor found 
by dividing the total quantity of alcohol by the quantity re¬ 
moved from the graduate, thus determining the total number 
for the particular level and series. In the heavier catches i 
of the total number was actually counted and the factor was 6. 
It would probably have been better had the quantity of alcohol 
oeen so chosen that the factor might have been 10. The selec¬ 
tion, however, was determined by the instruments at hand. In 
the catches containing but few specimens, as in the 12-15 m. 
and 15-18 m. levels, the entire quantity was counted. In the 
9-12 level the catch was diluted to either 16 or 12 cc., and from 
this quantity 4 or 6 ccm. respectively would be counted. 
After determining the number of the smaller Crustacea, the 
contents of the cylinder were poured into a shallow crystallizing 
dish, which was placed upon a black tile to facilitate observa¬ 
tion. All large types, such as Leptodora and the larvae of 
Corethra and Chironomus , were then counted directly. 
After the catch made during any of the periods had been 
counted, the cards were arranged in order, the results were 
posted, and the totals and percentages for the different levels 
were computed. A specimen of a part of such a sheet is given 
herewith. 
