4 SOUTH BUFFALO FLOODS AND PROPOSED REMEDY. 



Apparently the statement of flood flow for the Buffalo River 

 below the junction with Cazenovia Creek is less than for the 

 Croton River with a somewhat smaller watershed. This is not 

 inconsistent in the light of known conditions. Heavy rain storms 

 are greater and produce greater floods on the Atlantic slope of 

 the Alleghany Mountains than on the western slope. The same 

 may be noted of the Perkiomen and Ramapo floods. 



The flood of February 9th, 1900, reached a very high stage 

 in the lower portion of the Buffalo River, due to a westerly 

 gale, together with an ice and boat jam. The estimated dis- 

 charge of this flood was but 18000 cu. ft. per second, and the 

 high water in the lower river was evidently due to the peculiar 

 conditions noted. 



On February 28th and March 1st, 1902, there appeared a 

 flood which produced the largest measured discharge, and 

 reached the highest recorded general level of overflow on the 

 upper parts of the streams within the city limits. The actual 

 measurements of flow in the channel gave a total of 24300 cu. 

 ft. per second. 



It is the custom to establish observers at several different 

 points when floods are expected and generally it is easy to pre- 

 dict a flood from 5 to 20 hours before it arrives at the city line. 

 This last flood came a little sooner than was expected, and 

 some of the observers were not aware of the conditions until 

 the morning of March 1st: however, two fairly continuous sets 

 of gauge readings — one at the head of navigation and the other 

 near the center of the flooded area, at the junction of the 

 Buffalo River and Cazenovia Creek, were obtained, which, with 

 the records of the United States Engineer's recording gauge 

 near the mouth of the river, give a very fair idea of what 

 occurred. 



The slope of this flood and others is shown upon the accom- 

 panying diagram of flood heights. 



These records of gauge readings, together with float tests 

 made at different points on the stream, where fairly accurate 

 measurements of channel capacity can be made, enabled us to 

 estimate the amount of flow of water under different flood 

 conditions. 



Measurements of the channel before and after floods have 

 enabled us to determine, with some degree of accuracy, the 

 effects of the floods in the matter of deposit in the river bottom 



