J. Garard on Navigable Canals. " 
as a basis for the distribution of the lifts of any number of 
contiguous locks on a canal. 
Since the expense of water, from any basin, necessary for 
adouble passage of boats in the lock which joins that basin, 
is always proportionate to the lift of the lock, when, accord- 
ing to the supposition generally adopted, the draft of water 
of the ascending boat is equal to that of the boat which de- 
scends, it is evident that, if we continue the same supposi- 
tion, the reciprocal condition of a proper distribution of the 
locks consists in adapting their lifts to the quantity of water 
that the adjoining reservoir destined to supply water, can 
furnish without inconvenience. 
This principle established, let us admit that the highest 
lock of a canal is constructed in conformity to it, it is clear 
that if the canal, from this point downward to its lowest level, 
experienced no loss of water by evaporation, or filtration, all 
its locks should have the same dimensions as the first; for 
then the water drawn from the first level, would pass to the 
second, thence to the third, and soon successively to the 
last and lowest. 
In case of a negative expenditure, the same volume of 
water would ascend through all the locks in succession, from 
the lowest to the summit level.* 
Thus, whatever might be the number of locks, the de- 
scent of one boat and the ascent of another, throughout the 
whole length of the canal, would only require the positive or 
negative expense of water necessary for the double passage 
of those boats through any one of the locks. 
But the above supposition does not hold good in practice. 
Each successive level of a canal loses, necessarily, a certain 
quantity of water by evaporation, and is more or less ex- 
posed, according to the nature of the ground, to lose anoth- 
er portion by filtration ; therefore the volume of water drawn 
from an higher to a lower level by the first lock, cannot be 
* The expense of water can be negative only while the boat descends, and 
must always be positive when the boat ascends the lock; hence the nega- 
tive expenditure begins by flowing from the highest lock into the summit 
level, from the second lock into the first, from the third to the second, and so 
on to the last andlowest level. The positive expense begins at the upper lock, 
when caused by the descent ofa boat, and at the lower one when caused by 
the ascent, and follows in the same succession. The author of the Memoir 
hasalready explained this principle at the commencement, but has here fal- 
len inte an erroneous expression..— Translator. 
