THE MISSISSIPPI VALLEY. 47 



These dates are chiefly taken from the newspapers of that day — the papers 

 of each city generally recording the time when the water ceased to threaten 

 their respective towns with increasing calamities. From these facts it appears 

 that that remarkable wave travelled from Pittsburg to Cincinnati — a distance 

 computed at 460 miles — in 7^ days; or at the rate of 61 miles a day, or 2 ^ 

 miles an hour. The rate of the progress of this flood was, therefore, but a frac- 

 tion more than half as great as the surface velocity of the river in the channel 

 at Wheeling ; and but two-thirds of the velocity of the whole mass of water, 

 after it had attained the height of 35 feet. 



We do not know, from actual measurement, the precise velocity of the cen- 

 tral surface thread of the Ohio at the top of this flood ; but computing it from 

 velocities carefully observed up to 31^ feet, it must have been about six miles 

 an hour, or more than twice as great as the actual progress of the flood itself 



This apparent paradox is susceptible of easy and satisfactory explanation. 

 The flood coming from the mountains finds the channel nearly empty as it 

 approaches. The water which rolls past any given point has, therefore, a 

 double duty to perform ; first, to fill the charmel below, and then to supply the 

 volume which the channel below discharges. The wave is partly spent in filling 

 the empty channel, and is retarded while being replenished from above. 



The same circumstance which explains the retardation of the flood-wave, 

 accounts, also, for another fact, observable in the progress of all great freshets — 

 that, while the water is rising, the drift leaves the channel, and tends towards 

 the shores ; and as the surface falls it recedes from the shores, and seeks the 

 thread of the channel. The swiftest water is near the centre of the river, where 

 there is always the least resistance. The empty stream below is, therefore, first 

 supplied by the central fillet. The foremost water comes down along the centre 

 and at the surface, and flows from the centre towards the shores, to fill up the 

 vacant space. The river, while on the rise, is consequently higher at the centre 

 than at the borders ; and the drift descends on the surface, following the lateral 

 current towards the banks. 



But, as the water falls, the effect is reversed. The supply first fails above, 

 while the draught still continues from below, and is most rapid at the centre, 

 where there is still the least resistance. The void occasioned by the draught is 

 consequently first produced in the centre, and the water flows in from the 

 shores to supply the void. The drift again obeys the lateral current until it 

 I'eaches the channel, which, by this excess of the central draught, is depressed 

 below the surface at the shores. The drift floats down the slopes of this depres- 

 sion, and has no power to ascend the plane on either side to regain the shores. 



These considerations explain the difficulty, well known to pilots on this river, 

 of keeping the channel on a rising stream. The water being then highest in the 

 centre, the boat is continually drawn towards the banks ; while every raftsman 

 knows that it requires no skill to float down the current when the flood is sub- 

 siding. This fact is well expressed in the language of an old flat-boat man, 



