6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 134 
SOME RESULTS OF STUDIES OF WEATHER PERIODS 
“Sixty-year Weather Forecasts,” my Saint Louis precipitation 
paper, may be thought to have a sensational title. Critics are apt to 
say that when I use all the monthly mean values of precipitation at 
Saint Louis from 1854 to 1939 as a basis to determine the form and 
amplitude of 23 periods, no part of the synthesis of them between 
these dates is a forecast. On the contrary, 1,032 months are used to 
determine the features of these periods. No year has more than 12 
months. Hence the form and amplitude of the curve representing the 
march of precipitation in any one year between 1854 and 1939 can 
12 
have no more than 
10 
=1I.2 percent influence on that year’s curve. 
2 
Therefore, each year’s march of precipitation between 1854 and 1939 
is practically an independent forecast. After 1939, up to 1957 when 
my synthesis ends, all years are completely independent forecasts. As 
the halfway point between 1854 and 1939 is the year 1897, forecasts 
may be regarded as made from 1897. 
Thus, speaking approximately, every year from 1854 to 1957 is 
forecasted as if from 1897. As a fair specimen of such forecasts, I 
cite from “Sixty-year Weather Forecasts’: figure 1 (here fig. 1), a 
facsimile of the 5-year Saint Louis forecast, 1875-1879; figures 2, 
3, and 4 (here fig. 2), comparing forecasts with events for 6-year 
intervals, 1934-1939 in percentages of normal in the precipitation at 
Peoria and Saint Louis, and in the temperature departures from the 
normal at Washington, D. C.; and figure 5 (here fig. 3), comparing 
synthesis and event for Saint Louis precipitation, 1800-1887. Finally, 
I cite figure 6 (here fig. 4), comparing predictions for 1952-1957 of 
precipitation at Saint Louis and Peoria, prepared, of course, from 
wholly independent data. The predictions are in effect based on the 
year 1897, the halfway point between 1854 and 1939, which were the 
extremities of the basic interval employed. These two stations, Saint 
Louis and Peoria, are about 100 miles apart. Their 60-year forecasts 
run almost parallel. Both indicate the approach in 1952, waxing, maxi- 
mum severity in 1956, and probable end of the drought in 1957. 
The tabulation of Saint Louis precipitation for 104 years (1,248 
months) is preserved at my home in a roll 25 columns wide, 1,248 
lines long, and about 20 feet from end to end. Comparing its two 
curves of synthesis and event for 100 years, 1854-1954, 70 of the 
100 years were of the same degree of similarity in time, form, and 
amplitude of range, shown by figure 3 of the present paper. In the 
other 30 years the features of the parallel curves were similar, but 
