28 THE AMERICAN NATURALIST.  [Vov. XXXII. 
become changed into food material suitable for them; conse- 
quently, the organisms develop and the quality of the water 
suffers. 
This matter of stagnation with its unpleasant effects is an 
important one for water-works engineers to study. The growth 
of organisms is closely connected with the presence of organic 
matter at the bottom. Observations have shown that if the 
organic matter is absent the organisms do not grow to any great 
extent. The best modern practice in the construction of reser- 
voirs for water supply, therefore, indorses the removal of the 
soil from all areas to be flooded. This, however, is usually a 
matter of great expense, and, for that reason, some engineers 
do not consider it advisable to remove the soil from very deep 
reservoirs. In the opinion of the writer this position is well 
taken only in the case of reservoirs so situated and so deep that 
there is practically no circulation of the water at the bottom, 
and, therefore, no opportunity for any foul matter to be carried 
upwards. ; 
The requisite depth for the attainment of such a condition 
is at present unknown. We know that in some very deep lakes 
the water at the bottom remains constantly at the temperature 
of maximum density, but we do not know how much this depth 
must be diminished in order to have circulation take place. 
Moreover, the depth is not the only factor concerned. The size 
and shape of the lake, its geographical location, and the nature 
of the surrounding country all have their effect upon the 
circulation of the water. As the vertical circulation of water 
can be studied best by means of its temperature, we see how 
valuable it would be to have regular and continued temperature — 
observations made at various depths in our deep lakes and 
ponds. The observations thus far made are far too few to 
enable us to establish the point desired. 
According to the classification here suggested, lakes and 
ponds are divided into three types, according to their surface 
temperatures, and into three orders, according to their bottom 
temperatures. The resulting nine classes are shown in Fig. 2. 
On these diagrams the boundaries of the shaded areas represent 
the limits of the temperature fluctuations at different depths. 
