348 MISCELLANEOUS STUDIES 



In the same way the following values of the constants for the 

 latitude interval from 20 N to 30 N were determined : 



a\ = .343, o' 2 = .0159, a 3 = .0161, k.m^ 



2.= the latitude gradient = .41, 



.155 = the latitude gradient of the half range, 



5.535 .523 m, m, 



B=- -i and tan e -= -2^-^7^7 = 2.41 



I ll-i _ I I {/ 



Substituting these numerical values in equation (22) gives for the 

 surface temperature 



6 = (3.35 + .155*) cos (30* 67.5) + 21.30 .41* (40) 



Observed and theoretical lag of temperature maxima and minima 



behind the radiation maxima and minima. Comparison 



of computed and observed normal temperatures. 



The value of e = 69, corresponding to the latitude interval 30 N 

 to 40 N, was deduced from theory; and since 30 corresponds to 



69 



one month, 69 corresponds to =2.3 months, the theoretical lag 



oU 



,, ( maximum ) . , ,. , ,, ( maximum ) j. , 



of th - e j minimum f temperature behind the j minimum j radiation. 



The theoretical time of the maximum temperature is therefore 

 8.3, or about halfway between August and September; while the 

 theoretical time of the minimum temperature is 2.3, or about halfway 

 between February and March. According to Kriimmel (1907, p. 407) 

 from numerous and extended oceanic observations the average time 

 of the lowest temperature is February (t = 2) and that of the highest 

 is August (t = 8). 



Again, from the three curves (Makaroff, 1894, pi. 26) giving the 

 mean monthly surface temperature observed in the North Pacific 

 between latitudes 30 N and 45 N the minimum temperature occurred 

 when 



* = 1.8, 3.0 and 2.5 



and the maximum temperature occurred when 

 f = 8.1, 8.3 and 7.8 



respectively. The average of the above values of t is 2.4 for the 

 minimum and 8.1 for the maximum. Thus the predicted value of the 



