THE MONSOON CLIMATE OF ARIZONA AND NEW MEXICO. 



13 



the rainfall at high altitudes is influenced much less than at low. As yet, data obtained 

 on the subject are not sufficient to permit of any trustworthy conclusions. The matter 

 is mentioned here merely as one of the many interesting problems which would repay 

 investigation. 



Another problem of the same kind is illustrated in figures 2 and 3. The curve of 

 figure 2 represents the total rainfall by years from 1868 to 1912 as given in the Summary 

 of the Climatological Data for the United States and in the Monthly Weather Review. 

 Figure 3 shows the summer rainfall for the six months from May to October inclusive and 

 the winter rainfall for the sLx months from November to April during the same term of 

 years. The latter two curves seem to indicate a reciprocal relation of some sort between 

 the rainfall of summer and winter. In general, when the sununer rains increase in amount 

 the winter rains decrease, and vice versa. This phenomenon is not confined to Tucson, 

 but is apparently characteristic of the Southwest as a whole. For instance, in the two 

 curves at the lower left-hand corner of figure 11, on page 109, it is clearly seen in the rain- 

 fall of Flagstaff, 200 miles north of Tucson and 5,000 feet higher. Inasmuch as a similar 

 relation between the rainfall of equatorial and temperate regions has been inferred from 



Summer rainfall 

 Winter rainfall 



Fig. 3. — Winter and Summer Rainfall at Tucson, Arizona, 1868-1912. 

 = "Winter rainfall, Nov .-Apr. = Summer rainfall, M,iy-Oct. 



68 1870 72 74 76 78 1880 82 84 86 88 1890 92 94 96 98 1900 02 04 06 08 1910 



Inches 

 10 



Fig. 4.— Comparison of 3-year Means of Winter and Summer Rainfall at Tucson, Arizona, 1868-1912. 



the comparison of records in various parts of the world, particularly India, it is of great 

 interest to find it so clearly manifest here. Examination of the curves shows that in two 

 cases out of every three a minimum of winter rain is followed by a maximum during the 

 succeeding summer. One would expect to find the reverse also true, and that a summer 

 maximum would be followed by a winter minimum, but this does not hold good. A 

 summer minimum, however, is usually followed by a winter maximum. In other words, 

 if it be permissible to generahze on so small a basis of fact, the minima appear to be the 

 critical points. A maximum, either in summer or winter, is not hkely to be followed by 

 especially marked conditions in the succeeding season. A minimum, on the contrary, 

 whether in summer or winter, is likely to be followed immediately by a maximum in the 

 succeeding season. 



The preceding generalization obviously holds good only about two-thirds of the time. 

 In figure 4 the summer and winter curves have been smoothed by using 3-year means 

 instead of the actually observed rainfall. WTien the minor fluctuations are thus eliminated, 

 the opposed phases of the summer and winter curves are brought out clearly in the period 

 from 1868 to 1887, and less clearly from 1895 to 1907. In the period from 1888 to 1894 



