234 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. JO 



to solar changes, and therefore we have studied the published tem- 

 perature data of Arctowski (1912). Here we have collected data 

 for curves which show the variations of temperature amplitude 

 (in Fahrenheit degrees) in February, July, and for the whole year 

 in Arequipa, Peru (fig. 87). We find in February, which is in the 

 southern summer, a well-marked maximum about the year 1905, 

 but the stro'ngly-marked minimum in 1907 that is found in all 

 curves of the amplitude of this station does not fall in with the 

 sun spot curve or with the prominence curve which has its maxi- 

 mum in this year. 



If it should be objected that we are dealing here with another 

 doubling of the sun spot period, it must be taken into account that 

 this minimum in the year 1907 was considerably lower than the 

 minimum in the years 1900 to 1902, when the sun spot minimum 

 prevailed. At all events the agreement between the curves of tem- 

 perature amplitude and the sun spot curve at this station is not 

 good enough to base any considerable conclusion upon it. 



THE YEARLY AMPLITUDE OF THE TEMPERATURE IN NORTH AMERICA 



If great variations in the relation between the incoming radia- 

 tion and the outgoing radiation of the earth take place from year 

 to year, one would expect that these would be particularly noticeable 

 in the inner parts of great continents, where the difference between 

 winter and summer temperatures is great, since there the summer 

 temperature is more strongly influenced by the incoming radiation 

 and the winter temperature by the outgoing radiation. We have 

 therefore examined the yearly amplitude, that is to say, the tem- 

 perature difference between the warmest and coldest parts of the 

 year, in four different regions of the United States. The result 

 is given in curves I to IV of figure 88. It will be seen that the 

 fluctuation on the Pacific coast (the Pacific states curve I) is 

 considerably more regular than in the other three regions, and it 

 goes for the most part oppositely to these. At the bottom of the 

 figure are the curves S and P for sun spots and prominences, the 

 latter according to observations at Rome and Catania. The reader 

 will see that there is no distinct agreement between the four curves 

 of yearly amplitude and these curves. The two temperature curves, 

 II and III, show marked minima in the year 1890 which in curve 

 IV is displaced to 1891. This is simultaneous with the 'minimum 

 in the sun spots and the prominences. But on the other hand in 

 the years 1901 and 1902 there is no corresponding marked mini- 



