Chase. } 182 
[Aug. 18, 
The tendency to opposition between rainfall curves on opposite sides of 
the Atlantic, of which I have already submitted some illustrations to the 
Society (ante, p. 38, &e. ), is interestingly shown near the solstitial and 
equinoctial per iods, by Table IV.* Columns 6 and’ 7 of the same table 
indicate a similarity between the curves of daily and annual rainfall at 
Philadelphia, which lends additional interest to my comparison of pluvial 
and auroral curves (ante, pp. 121-22). 
EXPLANATION OF FIGurss. 
The horizontal lines represent the average rainfall ; each vertical space 
represents a deviation of .2 of the mean value ; each horizontal space 
represents two days in the abscissas of the lunar ‘curves, or +; of a year 
in the abscissas of the solar curves. The lunar curves begin and end with 
the day of new moon; the solar curves with January 1. 
FORE Cet R OR SEER EARRMBEGED 
EGAN EVONVs ONREES 
PAE fap HN ale 
| Ng 
PACELLI bled 
BABN 7970 81 /BREABBD 
ERLE Eo) 
cual 
Sea pL a! CUTPSGAEEGE HY PSUR 
EN 
Fig, 1.—Lunar Curves, 
December to March, inclusive : ; continuous line. 
April to July, inclusive ; broken line. 
August to November, inclusive ; dotted line, 
Solar Curves. 
1855, Rs ‘61, 64, ae 70; continuous line. 
1856 62, 165 5, ; broken line. 
1857, 60, 63, 66, 160: dotted line. 
Big 2. 
1855-60 ; continuous line. 
1861~65 ; broken line. 
1866~’70 ; dotted line, 
* ?—P 
ublished by permission of Prof, Benjamin Peirce, Supt. U. 8, Coast Survey. 
