32 THE AMERICAN NATURALIST. (VoL. XXXII. 
occur not in summer, but in winter. In the first order this cir- 
culation period would be brief or entirely wanting; in the second 
it would be of longer duration; in the third order the water 
would be liable to be in circulation the greater part of the year. 
Tropical lakes are quite numerous, but GUSEr VALID ns are lacking 
to place them in their proper order. 
Most of the lakes of the United States belong to the tem- 
perate type. In this type there are two periods of circulation 
and two periods of stagnation (except in the third order), as we 
have seen illustrated in the case of Lake Cochituate. In lakes 
of the first order the circulation periods would be very short or 
entirely wanting; in the second order the circulation periods 
would be of longer duration; in the third order the water would 
be in circulation throughout the year when the surface was not 
frozen. 
If we recapitulate in tabular form, we have the following: 
CIRCULATION PERIODS. 
PouLar Types. 
TEMPERATE TYPE. 
TRropicaL TYPE. 
Ist Order. 
One circulation 
period possible, 
in summer, but 
generally none. 
Two circulation 
periods possible, 
n spring and 
fall, but gener- 
ally none. 
One circulation 
— possible, 
n winter, but 
generally none. 
cai il a Neg a a 
2d Order. 
One circulation 
period, in sum- 
mer. 
Two circulation 
periods, in 
spring and 
autumn. 
One circulation 
perio in win 
ter 
3d Order. 
Circulation at all 
seasons, except 
when surface is 
Circulation at e 
frozen. 
Circulation at all 
seasons. 
Speaking in very general terms, we may say- that lakes of the first order have 
no circulation ; lakes of the third order have no stagnation (except in winter); 
and lakes of the second order have both circulation and stagnation. 
In view of the comparatively few series of observations of 
the temperature of our lakes, the writer refrains from making 
any classification of the lakes of the United States, but the 
