21 



to the ground is comparatively slight, owing to the slow conduc- 

 tivity of the soil and the small specific heat of the atmosphere. This 

 point has been carefully developed by Maurer, of Zurich (1885). 

 But when rain and snow fall, then the latent heat formerly con- 

 tained in the atmospheric vapor is qniclvly given to the surface soil 

 and directly conducted deeper into the ground, and the latter is 

 warmed or cooled according as the rain or snow is warmer or cooler 

 than it. In general, the warming of the soil by warm rain is less 

 important than the cooling by cold rains, melting snows, and evapo- 

 rating wine's, 



CLOUDINESS. 



AATien clouds intervene the soil receives a smaller proportion of 

 direct solar heat, and the proportion diminishes as the thickness of 

 the cloud layer increases or as the jjroportion of cloudy sky to clear 

 sky increases. We may adopt the approximate rule that the warm- 

 ing elfect of the sunshine is inversely as the cloudiness of the sky 

 within 45° of the zenith; thus for a sky covered by 10 cumulus or 

 10 stratus the direct solar heat at the ground is 0;, for 10 cirrus or 

 cirro-cumulus or cirro-stratus the solar heat is about 5, while for 

 cloudiness the radiation that the observer receives is 10. 



SOIL THERMOjMETERS. 



The motions of the clouds do not affect the sum total of the 

 intensity of the sunshine, but the variations of cloudiness are so 

 important that it is best to make use of some form of sunshine 

 recorder or, better still, some form of integrating actinometer as a 

 means of determining the relative effectiveness of the sunshine for 

 any hour or day. If any such instrument shows that during any 

 given hour, with the sun at a known altitude, the duration of the 

 effectiveness of the sunshine was the nth part of the maximum value 

 for clear sky, then we may assume that the heating effect of the sun 

 on the surface of the soil was the nth part of its maximum value 

 and may thus ascertain and, if need be, approximately compute the 

 irregularities of the diurnal waves of heat that penetrate the soil. 

 But these ii-regularities are directly shown by thermometers buried 

 in the soil at different dejjths, and the observation of such soil ther- 

 mometers is an essential item in the study of climate and vegetation. 

 The absence of these observations has necessitated much labor in 

 unsatisfactory efforts to obtain the approximate soil temperatures 

 from the ordinary observations of air temperature, radiation ther- 

 mometers, clouds and sunshine. 



Fortunately the agricultural experiment stations of the United 

 States have begim the observation of soil temperatures as- distin- 



