464 LIFE OF HORACE BENEDICT DE SAUSSURE 



case I am inclined to attribute the cold air to the downward flow 

 of air chilled by radiation or by contact with snow on mountain 

 slopes through scree material which prevents free mixture with 

 the atmosphere ; but I am not familiar with recent work on this 

 phenomenon, which may have another explanation. 



De Saussure also linked his observations on the temperature 

 of deep lakes with those on the temperature of the soil. 

 He made experiments for three years with thermometers at 

 depths of two to six feet, and also with similar thermometers 

 at the bottom of a pit about thirty feet deep. He found that 

 the seasonal range of temperature decreased and the period 

 was retarded as the depth increased, but on the average the 

 deeper he went the colder it was. He was aware of the fact 

 that deep-seated springs were usually warm and that in deep 

 mines the heat became very great ; but he attributed this heat to 

 chemical action, or, as he put it, ' fermentation,' and supposed it to 

 be a mere local disturbance of the normal fall of temperature 

 towards the centre of the Earth. He allowed, however, that more 

 observations were necessary to settle the question. He under- 

 stood that at a certain depth in the ground the effect of the seasons 

 disappeared ; but he did not see that, in order to ascertain the 

 real temperature of the Earth's crust, the observations must be 

 continued far beyond this depth. Had he known that the depth 

 to which the mean annual temperature of the soil continues to 

 diminish is greater in polar climates than in temperate, and in 

 temperate climates than in tropical, he might have been less 

 confident as to the probability of a cool interior. Throughout his 

 researches he seems to have missed the powerful effect of winter 

 cold alternating with summer heat. 



One obstacle to belief in the general fall of temperature with 

 depth below the surface was Donati's observation that the tempera- 

 ture of the deep parts of the Mediterranean was 10 R. at the 

 bottom. De Saussure had a special slow-action thermometer 

 made to test this statement. It had a thick glass bulb an inch 

 in diameter, was filled with spirits of wine, and the bulb was 

 embedded in a ball of soft wax more than six inches in diameter, 

 and the whole enclosed in an iron-bound wooden box. This 

 instrument required many hours before it took the temperature 

 of its surroundings ; but once it did so a long time elapsed before 

 the temperature changed by a perceptible fraction of a degree. On 

 two occasions off the coast of the Riviera this instrument was 

 anchored out for a night, once in 886, and once in 1800 feet of 

 water. Each time the temperature was found to be 10' 6 R. De 

 Saussure could not doubt the accuracy of his observations, which 



