394 
BULLETIN OF THE BUREAU OF FISHERIES. 
This suggestion carries no weight, because the origin of the "habit of swarming” 
remains unexplained. We know, however, that the diurnal migrations of many 
plankton Crustacea depend on light conditions, and Steuer (1910) has suggested 
that the gathering of plankton animals may be caused also by wind and current. 
There are probably various tropisms that cause the migrations of the plankton 
animals and their gathering on the surface or at a definite depth of the lake. The 
problem requires an experimental investigation, and a descriptive examination is 
insufficient to solve it. The explanation of the behavior of the animals, if one 
intends to explain it, ought to be based on exact data and not on purely speculative 
suggestions. 
LAKE PEPIN. 
Copepoda are very numerous in Lake Pepin and form a considerable part of 
its plankton. The mean number in this lake, computed from observations made at 
36 stations, reached 25,800 per cubic meter. In the Mississippi River, just above 
the head of the lake, the mean number, as computed from the results of the obser- 
vations made at three stations across the river, was 3,000 per cubic meter. Below 
the lake, at Reads Landing, their number averaged 20,000. If we take into con- 
sideration only the results obtained in the main channel of the river above and 
below the lake and omit the observations near the banks where the water runs 
slowly or is stagnant, we find that the water flowing into the lake carried about 
8.000 copepods in each cubic meter and that running out carried from 44,000 to 
46.000 per cubic meter (stations 96, 101, 104, Table 29, p. 429). The density of the 
copepod population in the different parts of the lake is shown in Figure 16, which 
is plotted in the same way as Figures 13 and 14. The figures on the lines and on 
the stations indicate the average number of Copepoda, in thousands of organisms, 
per cubic meter. 
It is easy to see that the Copepoda were more abundant in the lower part of the 
lake, where a large area with the maximum content of 70,000 per cubic meter 
could be found midway across the lake. The same number of Copepoda per cubic 
meter (about 70,000) occurred in the upper part of the lake, but the area was small. 
In the northern shallow part of the lake there were only a few copepods, their 
average ranging from 1,000 to 10,000 per cubic meter. 
The fluctuations in abundance of copepods from the head to the foot of the 
lake are shown in Figure 17. The figures on the ordinates indicate the average 
number of copepods for different cross sections of the lake, each figure representing 
the average of three stations across it. The first figure refers to stations 96 to 98 
(fig. 12, p. 383), located just above the head of the lake. The abscissae give the dis- 
tances in miles from this point. The results of the observations made in the north- 
ern shallow part are omitted. The increase of Copepoda in the lower part of the 
lake is clearly indicated. Their average frequence there is evidently greater than 
in the upper part, where only a local increase is found at stations 85 to 86. 
The distribution of copepods in general coincides with the distribution of the 
total amount of plankton in the lake, as is evident from Figure 17 and Table 21, 
and also by comparing Figures 12 and 16. 
