River Temperature. 309 
experiments on the temperature of streams, rivers, and 
ponds that may be a mile or so apart. For those who may 
wish to take the temperature at different depths, the Sixe’s 
instrument I employed, and which may be obtained from 
the same makers, is very serviceable. It also will supply 
the place of maximum and minimum thermometers, if 
checked by a standard instrument. 
With regard to the times and the frequency of the river 
observations, daily observations are of course the best, but 
those made every two or three days will give with fair 
approximation the monthly mean. The air observations 
should be made daily, and the maximum and minimum 
instruments should be read and set at 8 or 9 in the 
evening and noted for that day, together with the river 
observations of the same day. If the river’s temperature 
is taken late in the afternoon, an occasional sunrise observa- 
tion would give its diurnal range. And the more frequent 
the observations on the diurnal range, the more valuable will 
be the observer’s work. I have previously referred in detail 
to the hour of observation; the most convenient time will be 
probably when the river is at its maximum, usually in the 
middle of or late in the afternoon. 
One of the most important, and often the only available, 
means of comparing rivers, is to be found in the contrast of 
the curves of the monthly means for the whole year. As 
ably argued by Dr Ule, the difference between the air and 
the water means has a fairly constant value, less variable 
than, and therefore less influenced than the actual tempera- 
tures of the air and the water, by the changing climatic 
conditions from year to year. And from this point of view, 
an observer, who has never before interested himself in these 
matters, and whose period of observation may be limited, 
may be encouraged by knowing that even a single month’s 
work may afford material of great service. This paper would 
have lacked much without the observations made by Dr 
Marcet on the Nile and Dr Griffith on the Brahmaputra, 
though in both instances they cover only a few weeks. In 
fact, by comparing Dr Marcet’s results for March on the 
Lower Nile with those for the Lower Mississippi by Captain 
