INTERCHANGE OF HEAT VON BEZOLD 



413 



lation or smoothing, in that the curves show the greatest irregu- 

 larities in their upper portions. 



This irregularity is easily explained, since on account of the many 

 disturbances that we encounter just beneath the surface of the 

 ground, we can only expect fairly reliable mean values from many 

 years of observations made at short intervals of time. 



Since we have only monthly averages for Nukuss we must regard 

 it as purely accidental that among the tautochrones constructed 

 from these values there are any that are exactly perpendicular to 

 the surface of the earth and correspond therefore to the limiting 

 values of the energy stored up in the ground. Such cases would in 

 fact assume that the times at which these extremes occurred fell 

 quite near the middle of the two months respectively. 





FIG. 59 



If, however, we consider the values for the very highest strata 

 as too uncertain and fix our attention first on the curves from 0.4 

 meter downward we find then March and September as the months 

 of least and greatest heat stored in the ground. But it looks as if 

 the September curve does not correspond to the full maximum, 

 although in August the maximum is not yet attained. 



This seems to suggest that in Nukuss the increase of heat comes 

 to an end and the decrease begins before the autumnal equinox 

 (September 21), if we may be justified in drawing such a conclusion 

 from averages that represent so few years of observation. 



If now on the basis of these considerations we actually compute 

 the annual heat exchange for Munich and Nukuss by taking as a 

 base the earth temperatures of April 1st and September 28th at 

 Munich but the monthly means for March and September at Nukuss, 

 we find the following approximate maximum values of u 2 — u x 



For Munich 36 C m 

 For Nukuss 48 C n 



17 H. Wild: Ueber die Bodentemperatur in St. Petersburg und Nukuss. 

 Wild Repertorium, VI, No. 4, 1878, pp. 45, 46. 



