68 Messrs. W. N. Shaw and E. W. Cohen. On the Seasonal 



several maxima which occur at irregular intervals, showing that the 

 relative frequency of the two types of weather^ plays no part in causing 

 the second-order effect. Finally, by multiplying the percentage differ- 

 ence of frequency of cyclonic and anticyclonic weather for each month, 

 by the difference in temperature (excess of the cyclonic, in each case, 



DIAGRAM 2. 



EFFECT OF TYPE (CYCLONIC OR ANTICYCLONIC) 

 OF WEATHER ON TEMPERATURE. KEW 1876 TO 1880. 



DEC.JAK.FEB. MAR. APR.HW JUN. JUl. AUG. SEE OCT. NOV. DEC. 

 + 4 



fi'6. 1 Divergence from first order curve 

 Thick during cycLonic weather, 



Fjt>.. CycLonic temperature -AnticycLonic. 

 rm.<s. Frequency of cycLomc- Frequency ofanCicycLonic 

 Expressed AS percentage of tat&i number of days 



Fit>.4. Effect of the type <f we&ther on the. 

 divergence from first; order curve. 



being taken as positive, and vice versd), a resulting curve (fig. -1) is 

 obtained which gives the total effect of the type of weather on the 

 atmospheric temperature. A glance at this curve will show that this 

 effect is in no respect similar to the second-order effect. It is thus 

 evident that the cyclonic or anticyclonic type of the weather plays no 

 part in causing the second-order effect, which is nevertheless the only 

 prominent periodic temperature variation which has a meteorological 

 origin. 



