130 RELATION BORNE BY THE TEMPERATURE 



air begins to decline about the middle of July ; that of the earth 

 about the 1st of August; and as the latter does not rise so high 

 nor so quickly with a generally ascending temperature, so it 

 falls more slowly and not so low as the air when the general 

 temperature is declining. The diagram for the temperatures 

 at Trevandrum, exhibits two nearly parallel curved lines, the 

 earth averaging a little more than five and a half degrees higher 

 than the air. The greatest approximation is in August and 

 September, as is generally the case at other places in the tropics, 

 and likewise in colder latitudes ; but in the latter there is a 

 coincidence of the lines in March or April which is not followed 

 at Trevandrum. On the contrary, the powerful sun-heat 

 which there prevails in the months of February, March, and 

 April, appears to heat the earth more than the air, till the 

 setting in of the rainy season, in consequence of which the 

 earth is lowered from its maximum, 89"61 in AprU, to its 

 minimum 83'25 in July, or more than 6°; whilst the air is 

 lowered, during the same period, scarcely 5°. 



"From the foregoing facts and diagrams we can form a 

 tolerably correct idea of the relation which the monthly mean 

 temperature of the soil bears to the monthly mean temperature 

 of the air throughout the year. The latter is known at a vast 

 number of places ; but that of the earth comparatively at very 

 few. It may, however, be estimated with sufficient accuracy 

 for all practical purposes connected with horticulture and agri- 

 culture. For example, we may take the tables and diagram for 

 Chiswick as our guide, for all places having nearly the same 

 monthly and annual temperature. Where the winters are 

 colder, as at Copenhagen, we must add 2° or 3° more than 

 for Chiswick to the temperature of the air in January, in order 

 to obtain, approximately, the temperature of the soil for that 

 month. In AprU, throughout the world, from lat. 60° to lat. 30°, 

 the mean temperature of the earth and air may be considered 

 to average alike, or to differ rarely more than 1°. From April 

 to July, subtract about 2° from the monthly mean of the air 

 for that of the earth. In August the temperatures again coin- 

 cide. The earth, after this, maintains a higher temperature 

 than the air throughout the remaining months ; so that in 



