531 
data from the silt-laden planktons of Spoon River. The leak- 
age affects primarily the smaller flagellates, ciliates, and dia- 
toms. For reasons previously given, the leakage introduces 
proportionately a greater source of error in the Spoon River 
data, and, in so far as it is operative, tends to increase the con- 
trast between the two streams. From data now at hand it is 
not possible to estimate the extent of this error, but it should 
be borne in mind in considering the data in the table concern- 
ing these groups especially. In spite, however, of these sources 
of error, some points of contrast are so striking that certain 
conclusions seem justified. 
There is a marked contrast in the total population. The 
ratio of the average number of planktonts in a cubic meter 
of Spoon River water to that in a like quantity from the main 
stream is 750,429 to 28,283,295, or 1 to 88—a ratio much greater 
than that found in the volumetric comparisons. It is due to 
the relative preponderance of small planktonts, especially the 
limnetic diatoms, in the collections from the main stream, as 
will be seen on comparison of the ratios of the other groups 
listed in the table. If the low-water period of 1897 be excluded 
from both averages the contrast between the two streams will 
be very much heightened. The data from these grand averages 
of all the coincident collections in the two streams are thus 
strongly confirmatory of the contrast in productiveness dis- 
closed by the volumetric comparisons, and of the fact that 
the access of water from Spoon River causes a diminution in 
the number of organisms per cubic meter of water in the 
main stream. 
The comparison of the representations of the various 
groups in the collections tabulated, throws some light on the 
qualitative differences in the plankton of the two streams. 
The algae, other than diatoms, as in the case of all groups 
except the insect larve, are present in smaller numbers in 
Spoon River, the total ratio being 1 to 1.4—abnormally low be- 
cause of a single unusual collection in the autumn of 1897. 
The ratio in 1898 and the latter part of 1896, 1 to 10, is proba- 
