300 Proceedings of the Royal Physical Society. 
to determine this point. In summer the maximum is 
reached between 4 and 6 P.M., and retained up to between 
half-past 7 and half-past 8 in the evening, when the water 
begins to cool. In winter, as far as my data show, the temper- 
ature begins to fall after 2 p.m. In the spring and autumn 
the highest reading is obtained between 3 and 4 P.M, and in 
sunny weather the water does not begin to cool until 5 or 6. 
With regard to rivers of warmer climates, Dr Borius, in 
his table of the Senegal observations, makes it clear that the 
water reaches its highest temperature generally about 4 P.M. 
in August, the cooling between this hour and 8 P.M. being 
usually slight. From Mr Wallace’s observations in September 
on the Rio Negro, and from those of Dr Griffith in the same 
month on the Brahmaputra, it would appear probable that 
the former river is warmest between 3 and 4 in the 
afternoon, and the latter between 4 and 5. 
We are now in a position to form an idea of the suitable 
hour to determine a river’s maximum heat. For the temperate 
regions, in the case of a river of the size of the Thames 
or the Seine, 4 to 5 P.M. in summer, 3 to 4 in spring 
and autumn, and 2 in winter, would prove suitable hours, 
Smaller rivers, like the Loir at Vendéme, where it seems to 
be about half the size of the Thames at Kingston, should be 
observed an hour earlier during all the seasons except in 
winter, when 2 o’clock would doubtless do. Rivers much 
larger than the Thames and the Seine, such as the Danube 
and the Volga, would on @ priori grounds attain their highest 
temperature in summer not before 6 in the evening, and in 
the other seasons about 4 or 5 in spring and autumn, and 
2 or 3 P.M. inwinter. For great rivers in warm climates it is 
probable that the hour before sunset would afford the most 
fitting time for observation all the year through. 
With regard to the time when a river reaches the mean 
temperature of the day, Renou’s tables give important infor- 
mation for the Loir, and from them we may conclude that 
the mean temperature is attained at half-past 10 or 11 
AM. in summer, at half-past 11 in spring and probably 
autumn, and at noon in winter. For the Thames I very 
roughly estimated that the suitable time would be between 
