THE ENGLISH LAKE DISTRICT. 117 



In consequence of unforeseen difficulties, and the fatal occurrence above men- 

 tioned, I fear that all the observations obtained prior to July 1851 are open to 

 some degree of suspicion, and it is with no little regret that I feel compelled to 

 reject them in toto from the tables. 



In the observations made subsequently to July, 1851, every reliance may be 

 placed ; indeed, I could not have much greater confidence in their accuracy, if the 

 readings had all been taken by myself, as, about this time, I was fortunate in secur- 

 ing the services of a very careful and efficient registrar, who thoroughly understood 

 the work he had to attend to, and who gave entire satisfaction in the performance of 

 his duties up to the close of the year 1853, when the experiments were discon- 

 tinued. 



The mean difference between the absolute minima temperatures in the valley 

 at Seathwaite and on Sea Fell Pike was, in 1852, 13°-8, and, in 1853, 12°- 7, or an 

 average of 13°-2 in 2798 feet. This result, which may be assumed to represent the 

 average of the extreme difference at night in each month, is somewhat less than I 

 had anticipated, as it exceeds by a small fraction only the observed mean depression 

 during the day, derived from observations taken under ordinary circumstances. 

 In examining the tables of temperature for the Mountain Stations, it must be 

 borne in mind that it is impossible to ascertain the actual minimum temperature 

 of the air on the mountains in the winter months, as the boxes containing the 

 instruments (although raised four feet above the surface) are then generally 

 covered more or less thickly with snow. In the winters of 1852 and 1853, the 

 maximum depth of snow varied from 12 to 24 inches. 



The effect of snow in keeping warm the earth and objects upon it is well 

 known. Its slow conducting power is shown indirectly by the high relative tem- 

 perature indicated by the thermometer at Sprinkling Tarn, in January, February, 

 March, and April, 1853. The average difference between the thermometers on 

 Sea Fell and at Sprinkling Tarn is 5 ; but for the first four months of 1853, the 

 average difference was 12 -2. The quantity of snow deposited at the Tarn (pro- 

 bably from its more sheltered position) is invariably greater than on the Pike ; at 

 the middle and end of February, the excess was six inches, and by this extra 

 depth the temperature was raised 7 degrees. Mr Glaisher frequently found a 

 thermometer on the grass slightly covered by snow to read 8° or 9° higher than one 

 on grass clear of snow ; and, in one instance, when the temperature fell suddenly, 

 the difference (under three inches of snow) was no less than 34° ; and when the 

 thermometer on grass clear of snow had risen 30 \ (from -6° to +24°) no variation 

 had taken place in that under the snow, which still read 28°. Snow being so 

 perfect a non-conductor of heat, evidently prevents to a high degree the loss of 

 heat by radiation from bodies covered by it ; and it also prevents the loss of heat 

 from such bodies by conduction, at times when the temperature of the air is 

 lower than they are. We may therefore fairly assume, that when the self-regis- 



vol. xxi. part i. 2 I 



