36f> BULLETIN OF THE UNITED STATES FISH COMMISSION. 
The preceding figures were compiled from material in which two samples were miss- 
ing, but inasmuch as they were one surface and one bottom, a mid depth in the same 
vertical was used in the stead of each, and I believe the general results are very little 
disturbed. It here appears that on the return of daylight the organisms are again 
arranging themselves in the relations previously described, there being more total 
organisms at the surface than at either mid-depth or bottom, and while the total at 
bottom still exceeds those at mid-depth, this is due to the diatoms, which usually 
show this peculiarity of increasing at bottom, i. e., those here grouped as large species 
of Lauderia and Navicula. It is also noteworthy that in this series the copepods are 
not so abundant as they were in the preceding instance — i. e., the same section taken 
at low water at nighttime. The vertical at Station II also shows the most organisms 
on this series; and for all the stations the Melosira group is the most important factor, 
as it was likewise in the series taken eight hours earlier on this same date. 
I had fully expected to find a section taken at high water, or on a strongly flowing- 
tide, to be richer in these organisms than the same section taken at low water, but 
such has not proven to be the case. This is perhaps explained by the fact that as the 
longitudinal sections of Buzzards Bay show a marked decrease in this material as the 
open waters of the ocean are approached, so also a strong incoming tide in these 
shallow depths would tend to materially affect the numbers of organisms at the center 
of the bay by bringing in a great bulk of water from the outside, which is relatively poor 
in these. As this tide drifts out again the aggregate material is brought out from 
more inshore localities, thus increasing the amount of material in low-water analyses 
as compared with those taken at high water. 
In all that has thus far been proposed concerning the quantitative analyses of 
these organisms, the actual number per liter of ocean water has not been given, for 
the reason that the numbers tabulated were the ones actually observed under the 
microscope, from which the numbers in any given quantity of water must be esti- 
mated. The relative estimates here given are obtained in the following way: Five 
liters of water at each sample were filtered through a film of fine white sand at the 
bottom of a large funnel, the filtrate of organisms remaining upon the sand was then 
gently washed off in a small quantity of sea water and treated with a strong solution 
of formalin; when the material had thus been killed and had entirely settled to the 
bottom of the vial, the first formalin was decanted off and a fresh solution added, until 
the bulk of the formalin, including the filtrate, stood at just 15 cubic centimeters. 
Thus the organic material in a bulk of 5,000 cubic centimeters of water is collected 
into a bulk of 15 cubic centimeters of preservative, i. e., 333£ cubic centimeters of sea 
water are represented by every 1 cubic centimeter of the preserved material. The 
separation of the material — the filtrate — from the sand required the greatest care, but 
certainly all our errors were on the side of underestimation, inasmuch as we could 
not exaggerate the amount of organic matter in each 5,000 cubic centimeters of water 
used, and great pains were taken to transfer all the organisms from the sand to the 
preservative without loss. All the material here studied was collected in exactly the 
same manner by the same apparatus and persons. 
The next step in the estimation is to compute the number of organisms in one 
cubic centimeter of the preserved material, and this was done by means of the “Rafter 
cell” and micrometer (1 inch) eye-piece, the latter being so graduated into squares 
that one square in the eye-piece views a thousandth part of the surface of the 1 cubic 
centimeter, arranged on the stage of the microscope in the “cell”; for as the Rafter 
