294. Proceedings of the Royal Physical Society. 
Teddington, that both at the surface and at the bottom the 
temperature was 32°. On returning home I tested the 
correctness of my thermometer in a mixture of melting 
snow and ice. On another occasion, during this spell of 
Arctic winter, the water, as it rushed from under the ice 
down the weir at Teddington, was at 32°. Here it came 
within tidal influence, but the temperature remained the 
same, and no difference was to be observed at Petersham 
after the tide had been running up for two hours. My 
experiments were prolonged until late in the evening, and 
as I stood watching the river, with the air at 16°, and almost 
everything freezing around me except the water, there 
seemed to be in my surroundings a good deal to emphasise 
the opinion long since advanced by Weitz, that a river will 
not freeze solid until the soil is frozen to the depth of its 
bed. The literature on this subject is, however, already 
abundant, and a discussion would be foreign to the subject 
of this paper. 
To return now to the question of the surface-heating 
of rivers, I will assume that it has been well established 
that rivers of ordinary depth, say 2 to 5 fathoms, have 
either much the same temperature top and bottom, or show 
only a difference small in amount. No doubt the more 
rapid the current of a river, the more frequent and complete 
will be the mixing process, and a sluggish stream like the 
Thames will exhibit more surface-heating than a more 
impetuous river. Small streams, like the tributaries of the 
Thames, when in summer the water is low and the current 
very sluggish, present less uniformity in this respect. The 
Mole, which joins the Thames at Molesey, presents a typical 
instance. In the middle of May, about Hersham and Esher, 
in places where the current was scarcely perceptible and 
the depth from 5 to 9 feet, the difference between the 
surface and bottom temperatures varied from 2°°5 to 3°5 (F.). 
In those small rivers which, in dry summers, shrink con- 
siderably, the water becoming semi-stagnant in the deep 
' pools, the surface-heating would be great. Mr A. Harper 
gives a good example in the Thurso, where in summer the 
salmon die off in numbers in the heated waters of the slow- 
