126 



NEW YORK STATE MUSEUM 



annual runoff varies inversely as the size of the catchment, and 

 that maximum runoff, or flood flow, varies directly as the size of 

 the catchment. 



In order to gain some idea as to the applicability of this propo- 

 sition, the resume of discharge data, in the Twentieth Annual 

 Report of the United States Geological Survey, pages 46-64. has 

 been examined. This table includes about 225 streams in various 

 portions of the United States, with records ranging from 18 to 

 20 years in length to 1 year. A few of the best-known streams — 

 as, for instance, the Croton and Sudbury — are not given in detail, 

 although the large number included in this table, it is believed, 

 is sufficient to settle definitely this question. Only a very few of 

 the results will be referred to here. 



In the first place, it appears certain that with equal rainfall 

 there is no very definite relation between size of catchment area 

 and mean annual runoff. For instance, the Kennebec, at Water- 

 ville, Me., with a catchment area of 4410 square miles, has a mean 

 annual runoff for 6 years of 22.4 inches. The Cobbosseecontee, at 

 Gardner, Me., with a catchment area of 230 square miles, has a 

 mean annual runoff for 6 years of 18.5 inches. The Androscog- 

 gin, at Rumford Falls, Me., with a catchment area of 2220 square 

 miles, has for 6 years a mean annual runoff of 24.2 inches. The 

 Presumpscot, at Sebago Lake, Me., with a catchment of 470 square 

 miles, has a mean annual runoff for 11 years of 21 inches. The 

 Merrimac. at Lawrence, Mass., with a catchment area of 4553 

 square miles, has a mean annual runoff for 9 years of 21.3 inches. 

 Aside from the Androscoggin river these five streams support the 

 proposition that the runoff varies in some degree directly as the 

 catchment area instead of inversely. 



As to the maximum runoff, or flood flow, there is apparently 

 some slight relation, although even this is less definite than has 

 usually been assumed. 



As to the minimum runoff, there is apparently no relation, 

 extremely small flows happening on large streams as well as on 

 the smallest. There is, however, much more definitely a relation 



