Birge—Work of Wind in Warming a Lake. 
371 
DIAGRAMS. 
The following diagrams show in graphic form the chief 
results of this paper. Each figure has a somewhat full ex¬ 
planation and this need not be repeated in the text. In gen¬ 
eral it may be said that graphic methods, when they are 
applicable, are quite as accurate as computation by such 
methods as those which I have employed. Undoubtedly 
the curve of 1 — D in pi. VI can be drawn so that measurement 
will give more accurate results than computation by 1 m. 
intervals. If total results only are wanted the graphic method 
is also about as rapid as computation. But the gain in accu¬ 
racy is slight at best, and the computation is in all cases 
more accurate than the observations on which it is based, and 
far more accurate than the assumptions which accompany 
the observations. 
If the work done for or in the single strata is desired the 
results can be obtained much more quickly by computation 
than from diagrams. The measurement of numerous small 
areas consumes much time with little or no corresponding 
gain in accuracy. 
It will be noted that the longest computation—that by 
which the curve of distributed work is derived from the direct 
curve, can not be mechanically performed. 
