8 PHILOSOPHICAL SOCIETY OF WASHINGTON. 
The President remarked on the great interest of the discussion 
opened by Captain Durron’s communication to a preceding meet- 
ing, and expressed his especial approval of the method in which Mr. 
Warp had approached the subject. 
265TH MEETING. FrEeBRuARY 28, 1885. 
The President in the Chair. 
Fifty members and guests present. 
The Chair announced the election to membership of Messrs. 
Tuomas Corwin MENDENHALL, ALEXANDER ZiweEtT, Howarp 
Lincotn Hopextins, BAILEY WILLIs, JosEPH Paxson IppDINGs, 
and C. F. Marvin. 
Announcement was also made of the death of the Hon. Horace 
CAPRON. 
Mr. C. ABBE made a communication on 
METHODS OF VERIFYING WEATHER PREDICTIONS, 
giving a general account of the rules under which the U. S. Signal 
Office deduces from “indications ” and subsequent observations the 
published percentages of verifications. For purposes of prediction 
and verification the area of the United States is divided into a small 
number of districts. The “indication” for each district refers to the 
subsequent 24 hours, and is compared with the three next following 
weather-maps constructed from the observations, and the degree of 
correspondence for each station in the district is marked on a scale of 
five terms—0, 25, 50, 75, and 100. The published percentages are 
means of these marks. For certain special classes of phenomena— 
such as high-winds, frosts, and cold-waves—in which the indication 
only discriminates the occurrence and non-occurrence of a specific 
event, the formula for percentage of verification is 
v 
n +o 
in which n is the whole number of times the event is predicted, v is 
the number of verifications, or of events coincident with predictions, 
and o is the number of unpredicted events. 
It has been found for a large area in Europe, an area comparable 
