THE PANAMA ROUTE FOR A SHIP CANAL. 551 



that tbe total usc^s ot'watcj in the canal atWed to the loss by o\'apoTa- 

 tion, taken at <! in('li(\'=! in depth per month, from the surface of the 

 lake will amount to about 1,()T(» culjic feet per stn-ond if the traftic 

 through the canal should amount to 10, 000, 000 tons per annum in 

 ships of ordinary size. Tiiis draft per second is th(^ sum of -iot! cul)ic 

 feet per second for lockage, '207 for evaporation. ^."Xi for leakage at 

 lock gates, and '200 for power and other j)urposes, making a total of 

 1,063, which hsis l)een taken as 1,070 cul)ic feet per second. The 

 amount of storage in Lake Bohio ])etween the cdevations of 85 and 82 

 feet above sea level, as designed, is suthcient to suppl}" the needs of 

 that traffic in excess of the smallest I'ecorded low- water How of the 

 Chagres River during the dry season of a low-Vainfall year. The 

 lowest monthh' average tlow of the Chagres on record at Bohio' is 

 600 cubic feet per second for March, 181)1, and for the purposes of 

 this computation that minimum tlow has been supposed to continue 

 for three months. This includes a sensible margin of safet}^ In not 

 even the driest 3'(nir, therefore, can it b(^ reasonably expected that the 

 summit level of the canal would fall below the elevation of 82 feet 

 until the total traffic of the canal carried in ships of the present ordi- 

 nary size shall exceed Lo,0()0,000 tons. If the average sizt^ ot ships 

 continues to increase, as will prol)ably be the case, less water in pro- 

 portion to tonnage will l)e required for the purposes of lockage. This 

 follows from the fact that with a given tonnage the greater the capacity 

 of the ships tlu^ less the numlxn- recpiired, and consequently the less 

 will be the numl)er of lockages made. 



On the other hand it can be shown that Avitli a depth of 5 feet of 

 water on the crest of the (xigante spillway the discharge of that weir 

 2,000 feet long' will be at the rate of 78,2()0 cubic feet per second. 

 If the flood waters of the Chagres' shoidd flow into I^ake Bohio until the 

 head of water on the crest of the Gigante weir rises to 7i feet, the rate 

 of discharge over that weir would be 140,000 cubic feet per second, 

 which, as already shown, c^xceeds at least by a little the highest flood 

 rate on record. The operation of Lake Bohio as a flood controller or 

 regulator is therefore exceinlingly simple. Hie flood waters of the 

 Chagres would pour into the lake and inuuediiit(>ly begin to flow over 

 the Gigante weir, and contimie to do so at an inci'inising rate as the 

 flood continues. Th(> discharge of the \v<Mr is augnuMited by the 

 increasing flood, and decrease^ only at'tei' the passage of the crest of 

 the flood \va\'e. No Hood (n'cn as great as the greatest supposablo 

 flood on recoi'd can increas(> the elevation of the lake more than 02 to 

 92^ feet above sea level, and it will only \)o at long intervals of time 

 when floods will rais(! that elevation more tiian al)out 90 f^'et above 

 sea level. The control is automatic and unfailingly certain. It pre- 

 vents a1)solutely any damage from the highest supposable floods of 

 the Chagres, rfnd reserves in Lake Bohio all that is required for the 



