FIG. 10.—PERIODS OF RECORD OF STATIONS IN QUADRANGLES 12-B, 12-C, 12-D, AND 12-E. Fad 1870 1880 m n 1890 ་། Station Key Letters 1900 104 1910 1914 5420 FIG. 12. PERIODS OF RECORD OF STATIONS IN QUADRANGLES 12-J, 15-E, 15-G, AND 15-J. 1870 670 W X 1880 ab 1890 m Station Key Letters an ao 00000 agah 1900 precipitation, the quadrangles, dates of observations, and total precipitation at each station during such period. The card index so compiled includes a total of 2641 periods of excessive precipitation, which occurred during the years 1870 to 1914 inclusive. The cards were next separated into three classes, and each class was arranged in chronological order as follows: (a) Cards which contain only 1 station record, indicating storms with excessive precipitation areas so small that only 1 station was included. These cards number 1236 or nearly 50 per cent of the total. On account of the restricted storm areas which they represent no further use has yet been made of them. (b) Cards on which records from more than 1 but less than 6 stations appeared. These number 996 or over 35 per cent of the total, and represent comparatively local storms. (c) The remaining 409 cards, about 15 per cent of the total number, represent storms for which 6 or more stations reported excessive precipitation. They approach general storms in their characteristics, and to the heaviest of them extensive subsequent study was given, as described in the next chapter. Table 2 shows the number of storms in each class recorded for each year from 1871 to 1914. Table 2.-Total number of rainfalls causing excessive precipitation, recorded in eastern United States at 1 station, at 2 to 5 stations, and at 6 or more stations, from 1871 to 1914 inclusive. Table 2 shows that the total number of periods of excessive precipitation recorded yearly between 1874 and 1891 is fairly uniform. The years 1892, 1893, and 1894 show a tendency toward an increase and appear to be years of transition, and again from 1895 to 1914 the yearly totals are fairly uniform but appreciably greater than during the first named period. The effect of the number of observing stations on the number of excessive precipitation records is reflected in columns 4 and 5, but less so in columns 2 and 3. The explanation is that prior to 1892 the sparsity of stations was such that it was rare for even a large storm area to cause excessive precipitation at as many as 6 stations. This accounts not only for the small number of storms listed in column 4 prior to that year, and for the total absence of records of such storms during the years 1871, 1872, and 1873, but coincidently operated in placing many such storms in the 1-station and 2 or more station classes, thus swelling the figures in columns 2 and 3 out of normal proportion. It is therefore largely accidental that the number of storms in column 2 varies but little. It is fair to presume that the later figures in column 2 represent almost solely small area storms, such as thunderstorms, while the earlier figures include both this type and the storm covering larger areas. It appears to be satisfactory for purposes of comparison to separate the records into two 20-year periods; the first period embracing years from 1875 to 1894; the second period from 1895 to 1914. Table 3 shows the results of comparing the two groups. Table 3.-Comparative number of storms recorded in various classes east of the 103d meridian during the two 20-year periods 1875–1894 and 1895-1914. The conclusions which it is possible to draw from the card index of periods of excessive precipitation are somewhat meager. The anticipation was that the card index would furnish conclusive evidence of the relative sizes of different storms, and that it would place the excessive precipitation data in much more convenient shape for making frequency and seasonal studies. None of these objects was |