-1859] WRITINGS OF JOSEPH HENRY. 21 



summer by tardy exposure in the ground. This may per- 

 haps be most simply effected by burying a number of bot- 

 tles filled with water at different depths in the ground, say 

 one at the depth of 6 inches, another at 12, and a third at 

 18 inches. These in the course of time would take the 

 temperature of the earth in which they were embedded, and 

 would retain it sufficiently long unchanged, to admit of its 

 measurement, by inserting a thermometer into the mouth 

 of the bottle. 



No improvement is more necessary, for rendering the art 

 of agriculture precise, than the introduction into its pro- 

 cesses of the two essential principles of science, namely, 

 those of weight and of measure. All the processes in our 

 manufactories, on a great scale, which were formerly con- 

 ducted by mere guesses, as to heat and quantities, are now 

 subjected to rules, in which the measure of temperature and 

 the weight of materials are definitely ascertained by reliable 

 instruments. 



The foregoing are general views as to the great principles 

 which govern the peculiarities of climate, and especially 

 that of the United States, the truth of which, in reference to 

 our continent, and the modifications to which they are to 

 be subjected, are to be settled by observations in the future. 



In order however that the science of meteorolog}^ may 

 be founded on reliable data, and attain that rank which its 

 importance demands, it is necessary that extended systems 

 of co-operation should be established. In regard to climate, 

 no part of the world is isolated ; that of the smallest island in 

 the Pacific is governed by the general currents of the air and 

 of the waters of the ocean. To fully understand therefore 

 the causes which influence the climate of any one country, 

 or any one place, it will be necessary to study the condi- 

 tions, as to heat, moisture, and the movements of the air of 

 all others. It is evident also that, as far as possible, one 

 method should be adopted, and that instruments affording 

 the same indications, under the same conditions, should be 

 employed. 



It is true that, for determining the general changes of 



