502 SCIENTIFIC RECORD FOR 1883 



upper strata, and communicates at least a portion of their greater 

 velocity to the lower strata, or, as Buchan says, the influence of the 

 higher temperature of the earth's surface is "to tend to counteract to 

 some extent the retardation of the wind's velocity resulting from friction 

 and from the viscosity of the air." [This explanation loses much of its 

 force in cloudy weather, especially during extended heavy storms, when 

 the wind velocity still shows a marked diurnal periodicity, but this has 

 always seemed to the present writer explicable in consideration of the 

 fact that the solar heat is then all absorbed at the upper surface of the 

 cloud layer or of closely adjoining cumulus clouds and is doing its great 

 work there, i. e., expanding them and evaporating their vapor particles 

 and in both ways rendering them still lighter than the surrounding air; 

 their ascensive force is thus increased, the up-draft of air from the earth 

 the horizontal supplj winds, and the rotatory indraft, are increased, 

 and thus the diurnal period in our winds is maintained; similarly the 

 diurnal period in rainfall is maintained with a second maximum in the 

 early morning hours due to radiation from upper cloud surface and con- 

 sequent condensation of cloud particles into rain. Similarly the whole 

 development of an extended storm passes through its diurnal period, 

 including a periodic rise and fall of the barometric pressure due to the 

 varying force of the wind and resulting whirl; all which, as has often 

 been said, lies concealed in Ferrel's dynamical formulae.] 



The diurnal variation in the direction of the wind is illustrated by 

 the following figures, whence we see that even at sea the influence of 

 the sun in heating up neighboring lands or air masses makes itself visi- 

 bly felt in this item : 



Mean wind direction. 



Challenger, 1873. 



2 a. M. to 6 A. M 

 10 A. M. to 2 P. M 



In northeast trade 

 regions. 



E. 47° N. 

 E. 56° N. 



Mauritius observa- 

 tory. 



4 a. M. 

 1 P. M. 



E. 22° S. 

 E. 7° S. 



The percentage of sky covered with cloud is given in the last column 

 of the table of results of observations in the Challenger; the slight vari- 

 ations do not seem to the present writer to exceed the probable uncer- 

 tainty of the observations. The diurnal and annual variations in the 

 times of the occurrence of thunder-storms, i. e., the hours in which 

 thunder is heard, and which is, therefore, equivalent to the hours in 

 which lightning occurs, and is quite distiuct from the attending rain, 

 hail, or wind, is beautifully illustrated in the following table based on 

 fourteen years (1859 to 1872) of observations at Ekaterineburg, Ural 

 Mountains, which is given by Buchan without quoting the source (prob- 



