Conditions favorable to Glaciation. 



101 



Beg. F. 



Lowest January Latitude. 



Highest July Latitude. 



Isotherm. 



Asia. N. America. 



Asia. 



N. 



America. 



-50° 



60° 









— 40 



63 









— 30 



59 64 









— 20 



55 56 









— 10 



46 52 













45 49 









+ 10 



42 47 









20 



38 42 









30 



36 39 









40 



27 34 



77 









50 



25 31 



72 





69 



00 



21 



68 





66 



TO 



17 



59 





56 



80 



10 



48 





46 



90 







39 





43 



Some of these temperatures are affected to a greater or less 

 extent by the neighborhood of water as may be seen from the 

 charts. Thus the polar sea must certainly have a cooling effect in 

 Siberia in summer at latitude 77, and 40° must be there somewhat 

 too low a reading for a continental climate in July. Similarly 

 in winter 70° and 80° must be rather higher than the isotherms 

 in winter would be in southern Asia if land continued south- 

 ward. Again the great elevation of Central Asia and of the 

 Ttoeky Mountain region must give summer readings lower 

 than would be recorded were these areas at sea-level. 



Imperfect as these data are, they seem the best available, 

 and as such they are entered in diagram 2, an arbitrary unit 

 being selected as representing a Fahrenheit degree. Each X 

 represents an Asian observation, and each Y a North Ameri- 

 can one. 



It will be observed that the points representing temperatures 

 lie remarkably close to the curves representing the present rate 

 of receipt of solar energy, and thus seem strongly confirma- 

 tory of the trial hypothesis which, it may be well to state, was 

 not framed to fit the temperature observations. A still closer 

 agreement would be obtained by using a somewhat larger unit 

 for temperature. The climate represented by the curves is 

 intentionally shown as somewhat more extreme than the ob- 

 served climate because such would be the case if the air were 

 immobile. It might be easy to lay too much stress on the 

 close agreement of the curves and temperatures. I desire to 

 draw but one conclusion from it, viz : the mean rate at which 

 sunshine is received per unit area in continental climates is so 

 important an element in determining the seasonal deviation 



