REVIEWS 93 
affluents flow over gypsum, or whose shores are partly of gypsum. On 
the other hand, sulphuric acid is found in small quantity in lakes 
where there is no gypsum adjacent. Again, calcium is abundant in 
basalt, but not in the sediment in lakes in basaltic regions. Alka- 
lis are plentiful in granitic rock, but only sparingly present in the 
lakes in granitic basins. The alkalis and alkaline earths are carried 
off in solution, chiefly as carbonates, while the silica stays behind and 
is thus concentrated. (c) The composition of the sediment varies in 
different parts of the same lake. 
Sediment is absent in the bottom of the lakes where the slope is 
too steep for it to rest,in general where the slope is over 45°, and 
where local conditions have prevented deposition, as where springs 
enter or where drainage flows out. Sediment is also absent where cur- 
rents have been effective at the bottom. 
VI. Supply and loss.—A chapter is devoted to the supply and loss 
of water, and to the variations in the levels of the lakes. An interest- 
ing section is given to the average length of time which water stays in 
lakes. This is determined from the volume of the lake, and the rate 
of outflow. Thus in Lake Geneva, it is found that the average stay of 
water in the lake is eleven years and seventy-three days; in lake 
d’Annecy three years and one hundred and thirteen days; in Lake 
Chaillexon five days. Many other calculations are given, all of which 
tend to show that the duration of the stay of water in lakes is extremely 
variable. Lakes with surface outlets are found to change their levels 
but slightly. Data on this point seem somewhat imperfect, but the 
maximum known fluctuation in the case of lakes having surface outlets 
is three meters. In lakes having sub-surface outlets, fluctuations of 
level are far greater. They appear also to be greater for small lakes 
than for large ones. Thus the level of Lake Chaillexon between 
August 19, 1892 and December 31, 1895, fluctuated sixteen meters. 
VII. Zemperature.—The tables of temperature given show that the 
water at the bottoms of the deep lakes varies very little, and that it is 
near the temperature of greatest density all the time. The tables 
show that in most lakes there is a well-defined zone which separates the 
warm (during most of the year) water above from the cold water below, the 
transition being usually rather abrupt. This zone of transition is rarely 
more than twenty meters below the surface, and sometimes not more than 
ten. The causes determining temperature are considered. Aside from 
(2) climate, the effect of which is obvious, (4) the average depth, (c) 
