344 On Ground Ice found in the Bed^ of Rmminy Streams, 



formation of ground ice, that in the Teviot he often found it in 

 plenty in localities where the stream appeared to be well -shaded 

 by trees. The fact of ground ice being found in surh phui > ai once 

 snows that it never coukl have been produced there, lur to ioraa 

 surface-ice in small quantity under an arch requires a more intense 

 frost than we had had. The temperatux'e of the stream where 

 the ice was found under the arch was 32°, atmosphere 86°* 5 5 

 the thickness of the ice on exposed ditches measured a little under 

 2 inches. , ,l 



Wherever there are running streams readily cooled down to 

 32°, there can be no doubt about ground ice being found, but 

 this temperature is essential ; for if the water of the stream rise 

 only a small portion of a degree above the freezing-point, the 

 ground ice soon disappears. 



While examining the waters in the vicinity of Altcar, I have 

 often tried to estimate the relative power of a stream and of a stag- 

 nant water in producing ice. The Brook appears to come so soon 

 to 32°, that I incline to the opinion that more ice is made on it 

 than on an adjacent surface of still water. In the present frost 

 there appeared much less ice on it, but then the sun at this season 

 is daily of sufficient power to loosen large quantities from the 

 edges which are seen continually passing down to the sea. This 

 process has the effect of exposing every evening a fresh surface 

 of water to the most favourable action of the sky, where ice must 

 be formed more rapidly than in still waters already covere4iRV(ps 

 with thick ice. 



Subaqueous ice owes its preservation to the current of ice-cold 

 water which the stream keeps continually pouring upon it. On 

 the occasions when I have found it in the Brook, the frost had not 

 penetrated the soil or stagnant water deeper than 2 inches ; ice 

 could not at such a time remain in still water, at 2 feet below 

 the surface. In the situation I have described, where the sub- 

 aqueous ice had a covering of a stone arch in addition to the 

 water above it, the necessity of an active current of water to pre- 

 vent its melting was still more apparent. During a former 

 season, when examining the ice in the bed of the Brook, I tried 

 the temperature of the clay and sand over which the water roUedj 

 and found it to agree very closely with that of the water. Thisy 

 however, was only for a small depth, for during frosts the tem»^' 

 perature increased rapidly downwards for the first few feet. A 

 running stream is to some extent an exceptional case to the rule, 

 that in frosty weather the temperature increases quickly as we 

 descend ; for in the stream the force of gravity acts so as to 

 keep bottom and surface-water continually changing their respect- 

 ive places. On this peculiarity of a river must mainly rest the 

 view which I have advocated, that ground ice consists of spicula 



