76 Trans. Acad. Sci. of St. Louis. 



States for successive years, the values shown in a table 

 at tlie close of this paper and represented by the curve 

 T. It will be seen that the values are remarkably con- 

 stant, the highest in the fourteen years being 54°.4 in 

 1900 and the lowest 51°.4 in 1894, a difference of only 3°. 

 The mean for the period being 52°.9, the gTeatest varia- 

 tion was only 1°.5 above or below this. 



AVe m.ay say then, that taking the fourteen years as a 

 basis, the average annual temperature of the United 

 States, excluding the outlying parts, is 52°.9 F., and the 

 annual precipitation is 1,308 cubic miles. 



The state having the lowest amount of precipitation in 

 any year was Arizona in 1894, 5.8 inches; the greatest, 

 Alabama in 1900, 71.6 inches. The state having the low- 

 est average temperature was North Dakota in 1893, 35°.5; 

 the highest, Florida in 1897, 71°.8. These of course are 

 far from representing the extremes for small areas. The 

 maximum rainfall for single stations is not in Alabama 

 at all, but on the North Pacific coast, where in Washing- 

 ton and Oregon the rainfall is very often more than 100 

 inches per year, while some areas are of course practical- 

 ly rainless. 



Many attempts have been made to show a periodic 

 variation of the temperature and rainfall, and to connect 

 this jDcriod with some celestial phenomenon, as sunspots. 

 The chart shows both temperature and precipitation 

 curves together, and also the curve of Wolff's sunspot 

 numbers. While there seems to be a tendency, esiDCcially 

 in the first part of the fourteen years, for a minimum tem- 

 perature and rainfall to occur at a maximum of sunspots, 

 the latter part of the period covered is erratic in both 

 temperature and precipitation curves. The fluctuation 

 is a large fraction of the general periodic change which 

 coincides fairly well with the sunspot period. A contin- 

 uation of this work through the next sunspot period may 

 yield more conclusive results. 



Briickner has constructed a table of world tempera- 

 tures and precipitation from about 1731 to 1885, which 



