CLIMATOLOGICAL AVERAGES VON BEZOLD 417 



naturally belongs to them, independent of course of the uncertainty 

 that may affect individual numbers. 



'•'Correct average values can now be deduced at once by simple 

 mechanical quadratures from the data of the table or from the 

 ordinates." 



In the above-quoted memoir I have already referred to these 

 properties of the method hitherto employed and of the one here 

 recommended, except as to this last-mentioned point. 



This idea will now be further developed and applied to various 

 meteorological elements, and it will be shown how simply the con- 

 nection between the corresponding average values can be per- 

 ceived and what special considerations come to light almost spon- 

 taneously. 



This much being premised I now give the annual average values 

 of insolation, temperature and pressure of the air, cloudiness and 

 precipitation, first in tabular and then in graphic form arranged 

 according to the sine of the geographic latitude. 



As fundamental data I use the average values given in the 

 ordinary way [for degrees of latitude]; for the insolation I use the 

 values computed by Meech ; 2 for the temperature of the air I use 

 those given by Spitaler and Batchelder; 3 for the atmospheric 

 pressure, the numbers given by W. Ferrel ; for the amount of precipi- 

 tation, the figures given by John Murray, and finally for the cloudi- 

 ness, those given by Svante Arrhenius deduced from the charts of 

 Teisserenc de Bort; all of which are found collected in Hann's 

 "Klimatologie, " p. 217 [or Ward's translation, p. 100]. 



From these values by very careful graphic interpolation the 

 values were deduced that correspond to the series of values 0.05, 

 0.10 0.95 of the sine of the latitude. 



The values thus 'obtained are found collected in table 1 and repre- 

 sented by curves in fig 60. On the other hand, fig. 61 gives curves 

 whose ordinates are the arithmetical averages of the pairs of 

 values belonging to equal north and south latitudes. These latter 

 averages I call "holospherical" to avoid any misunderstanding, 

 while the two values belonging to each definite circle of latitude I 

 call "hemispherical." I will refer to this point in a subsequent 

 paragraph. 



2 See Hann : Handbuch der Klimatologie; 2d edition, vol 1, p. 103. [See 

 also L. W. Meech: "On the Relative Intensity of the Heat and Light of the 

 Sun." Smithsonian Contributions, IX, Washington 1857, or R. de C. Ward's 

 translation of Hann's Handbook of Climatology, 1903, p. 100]. 



' Hann: Handbuch, p. 200 [or Ward's translation]. 



