WEATHER BUREAU RIVER AND FLOOD SYSTEM 303 
and its tributaries, the area and topography of the catchment 
basin, the freqnenc 3 ^ and. especially the intensity, of the rainfall, 
the average time occupied in the passage of flood-crests from one 
station to anotlier, and the history of past rises. The knowl- 
edge of low-water conditions, especially where bars and shoals 
exist, is perha{)s of as great importance as the knowledge of high 
water. In fact, many statements are received at the central 
office in Washington from steamboat and navigation companies 
to the effect that low-water conditions continue longer and affect 
navigation more than those of high water. The people living in 
regions contiguous to navigable streams are materiall}^ affected 
in their industries by the conditions of navigation, but the de- 
struction of life and property, as effected by the rivers, depends 
entire!}’' upon flood conditions. The official in charge of a river 
center is expected, with the data at his command, to give in- 
formation to those interested in navigation, even during low or 
medium stages of water, that is of great pecuniary value; but 
his chief and foremost duty is the dissemination of warnings 
when floods threaten. 
Many data in regard to river stages have been published by 
the Weather Bureau, the Mississippi River commission, and the 
U. S. Signal Service. From the data thus collected and now 
covering many years at some stations and shorter periods at 
others, the following general relations have been deduced : The 
time it takes high water to pass from Pittsburg to Wheeling is 
one day; from Pittsburg to Parkersburg, two days; from Park- 
ensburg to Cincinnati, three days; from Cincinnati to Cairo, six 
days ; from Cairo to Vicksburg, seven days, and from Vicksburg 
to New Orleans, four days. From Pittsburg, therefore, to the 
Gulf requires 22 days. Similar general relations concerning the 
movements of other rivers have been determined. Since the 
time of travel is so great, it naturally follows that many inter- 
fering conditions arise tending to accelerate or retard the crest 
of the flood-wave. No simple time rules are therefore possible. 
The volume of water passing a station in a given time is known 
ftw only a few places, and varies, of course, with high and low 
water; nor can sini])l(.‘ rules be based upon the rainfall, as the 
absorptive condition of the soil is not constant and the distribu- 
tion of precipitation over the drainage area is not always deter- 
minable. 
'I'be principal rivers concerned in the W'eather Bureau sys- 
tem are the Alleghany, Monongabela, Ohio, Kanawha, W’abasb, 
