3 



number of species oi po ik i lotherms shows an East-West maximum (at roughly the Carpathians and 

 the Caucasus on one side, and the far East on the other side). These observations may be con- 

 sidered to be evidence that longitudinal distribution is dependent in Eurasia first and foremost 

 upon historical factors (the glacial impoverishment of the Siberian fauna). 



The Influence of CI i ma toloq i cal Eactors 



The primary role of temperature arid humidity amongst the physical factors is generally accepted. 

 Unfortunately, it is difficult at this time to obtain satisfactory information on many parameters. 

 Therefore, as a first approximation, three mean long-term quantities were used: air temperature 

 in January (ti), air temperature in July (t 7 ), and the sum of yearly precipitation in centimeters 

 (S). ' 



Analyses of variance using January temperatures gave the values presented in Table 5- A 

 weak correlation is demonstrated for all four taxa, but comparisons of the taxa do not lead to 

 clear-cut conclusions. The insignificance of the observed correlations makes it useless to 

 calculate empirical lines of regression (Table 6), but it should be noted that in all cases the 

 richness of fauna increases with increasing temperature. It is interesting also that the number 

 of species of reptiles and amphihians is almost exactly equal at low temperatures, but that, at 

 higher temperatures, the number of spec.es of reptiles rapidly exceeds the number of species of 

 amph i b lans. 



Table 5- Analysis of degree of influence of 

 the January temperature on numbers of species. 



M A R Am 



Table 6. The mean number of species found in 

 are as with different mean January temperatures 



t, in °C M A R Am 



r-1 4 A 



n-r 61 61 



N 2 0.42 0.37 



«JF 2 7006 43600 



646 



11 



4850 



An entirely different picture is obtained from the analyses of variance us'ng mean Juiy 

 temperatures (Table 7) : a very strong dependence is obtained tor all taxa except amphibians, 

 exceeding everything hitherto reported in this paper. tmpir'cal lines of regression are given 

 in Table 8. The dependence for mammals, birds and reptiles ma/ be descr i bed w i t h the following 



curvilinear equations: 

 Mammals: 



0.2052 + 0.027276x - 0.0026l2x 



O.IO94 - 0.0048x + 0.000185x 



Rept iles: 



13.0256 - 0.9828x + 0.0192x" 



the graphs (Figures 3, 4, 5) indicates sa t i sf act ory corr»a ondence be tween these curves and 

 the empirical data. The question of the amphibians will be considered below. 



Analyses of variance using the sum of the yearly precipitation gave the results presented 

 in lable 9. It is seen that the total precipitation plays, even as was expected, a more significant 



