22 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 60 



C is a combination of A and B. They are taken in the following 

 proportions : Multiply the percentage departure of radiation shown 

 in A by 6 1 and subtract from it the sun-spot number for the given 

 year. Curve D represents the departures of mean maximum tempera- 

 ture for 15 stations of the United States distributed all over the 

 country. It is smoothed in the same manner as curve A. Curve E 

 represents the departures of temperature for the whole world, also 

 smoothed in the same manner as curves A and D. The data for 

 the curves D and E are taken from Annals, Astrophysical Observa- 

 tory, Vol. 2, p. 192, and from the Monthly Weather Review. 



Although there is a considerable degree of correspondence between 

 curve B and curve D yet it is not hard to see that there is also much 

 discordance. For example, the sun-spot maximum of 1893 was 

 greater than that of 1883 or 1906, yet the temperature curve D 

 indicates a gradual increase of temperature for the three periods. 

 Also the temperature had begun to fall in 1890, although sun-spots 

 were still at the minimum; and the temperature had begun to rise 

 in 1892, although sun-spots had not yet reached their maximum. 

 Similar discrepancies occur in other parts of the curves. 



When, however, we compare the curves C and D, that is to say, 

 the combination of the effects of sun-spots and volcanic haze with 

 the mean maximum temperature for the United States, the corre- 

 spondence of the curve is most striking. It seems to us in considera- 

 tion of this, that there can be little question that the volcanic haze 

 has very appreciably influenced the march of temperature in the 

 United States. When we take the march of temperature for the 

 whole world the correspondence, though traceable, is not so striking ; 

 but in this case there are so many conflicting influences at work that 

 it is perhaps too much to expect so good an agreement. 



In view of this slight preliminary study of temperatures, it seems 

 to us that the question of the effect of volcanic haze on terrestrial 

 temperature is well worth serious consideration. Although a large 

 group of stations may by their contrary local influences mask the 

 influence of the haze, we believe it may be found eventually that 

 temperatures are influenced, perhaps as much as several degrees, 

 by great periods of haziness such as those produced by the volcanoes 

 of 1883, 1888 and 1912. 



1 Perhaps a better result would have come if 5 instead of 6 had been used. 



