2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 89 



solar periods have become available. The year maps from 1874 to 1882 

 do not extend west of the looth meridian ; the series of storm frequency 

 maps here published begins, therefore, with 1883, the first year that the 

 entire area of the United States was covered. Furthermore, according 

 to the Weather Bureau, the earlier maps are not statistically com- 

 parable with the later ones. A more extended network of stations 

 began with the year 1891. Although the year maps are here published 

 for the first time, the material for 1883-1912, assembled by months 

 in 1 0-year periods, was presented at the Second Pan American Scien- 

 tific Congress in 1915/ It is not the purpose of this paper to discuss 

 the storm frequency of the United States ; the maps, made with care 

 according to a uniform technique, are published in order to make the 

 material available to other investigators. How complex the factors 

 involved must be, will be evident from a comparison of the two years 

 1900 and 1911, both at solar minimum; a similar pair, also separated 

 by II years, 2 years after solar maximum, is 1919 and 1930. 



THE LATITUDE SHIFT 



With these two striking pairs in mind, it is evident that any uni- 

 form latitude shift within the 11 -year solar period must be masked by 

 other meteorological factors. The method of yearly departures from 

 a mean map suggests itself ; I have in manuscript a complete series 

 of such departure maps, which confirm the latitude shift, but illustrate 

 primarily the complexity of the factors involved. The most powerful 

 method of attack seemed to be to add together three years at solar 

 maximum and three years at solar minimum, and compare one set of 

 figures with the other. I have done this for five solar periods : 



Maximum 1882-84 versus Minimum 1877-79 

 Maximum 1892-94 versus Minimum 1888-90 

 Maximum 1905-07 versus Minimum 1900-02 

 Maximum 1916-18 versus Minimum 1911-13 

 Maximum 1927-29 versus Minimum 1922-24 



and present the results in Figures 2-6.^ This method allows for pos- 

 sible difiference in recording the tracks of barometric depressions, since 

 the comparisons are with periods separated by only five or six years. 



^ Monthly storm frequency in the United States. Proc. Second Pan Amer. 

 Sci. Congr., Sec. 2, vol. 2, pp. 338-391, 1917. 



^ Preliminary publication of parts of this study appeared in : Huntington, 

 The solar hypothesis of climatic changes. Bull. Geol. Soc. Amer., vol. 25, 

 PP- 477-590, 1914- Huntington, Earth and sun, New Haven, Yale Univ. Press, 

 1923. 



