788 MECHANICS. 
liable to the inconveniences and limited application of the 
fleam-engine alone. The agent he ufes, in addition to 
the fleam-engine, is comprejjcd. air ; and we think that an 
account of his invention will form a proper fuppleinent 
to the hiltory of fleam-boats. 
“ The method which I have to propofe has the advan¬ 
tage of requiring no wheels or other machinery to pro¬ 
ject into the water; it is obvious, therefore, that no more 
danger can be occafioned by it in ttormy weather to the 
vefl'el in which it is ufed, than in one navigated by fails 
alone. 
“ Having heard of various methods for impelling vef- 
fels, by the action of water forced in a jet againlt the me¬ 
dium in which the veflels floated, none of which appeared 
to me to be of much ufe, and forrie tried on a large fcale 
had no fuccefs, it occurred to me that the aftion o fair 
on the water would be more likely to fucceed : and, having 
made feveral experiments on the aftion of inclined planes 
in impelling boats, at a time w hen I refided near a large 
lake wftl calculated for fuch experiments, the mode which 
I have to defcribe, as well as I can remember, was lug- 
gefted by the recolleftion of thofe experiments applied to 
the new combination mentioned. In p>urfuance of this 
thought, I fil'd (in the year 1808) tried the operation of 
air forced into water through a fmall glafs tube ; and 
found that no effort I could ufe in blowing through the 
tube could force the air any perceptible diftance beyond 
its extremity, but that it rofe in bubbles, perpendicularly, 
from the tube direftly, in all cafes that I could command. 
1 then made fome fmall models of boats, with bottoms 
formed into inclined planes, fo inclofed by fide-fences, 
that air forced beneath them could only pafs in the direc¬ 
tion of the Item ; and I found that the air, rifing verti¬ 
cally from the end of the tube beneath the inclined plane, 
impelled the mode! with coniiderable velocity. The in¬ 
clined plane was more than an inch wide, and the tube 
was lefs than the fixteenth of an inch in diameter; pro¬ 
portions which may evidently be imitated on a larger 
fcale without danger of failure, fince the larger the tube 
the lefs proportional force, will be required to drive air 
through it, as its fides, which caufe the refiftance, bear a 
lefs proportion to its area; the fides of tubes being to 
each other as their diameters, and their contents as the 
fquares of thofe diameters. I had before experienced that 
the air forced from the end of the tube loit all velocity 
forward the inftant it parted from the tube as above men-" 
tioned, fo that the horizontal itnpulfe of the air could 
have no influence in moving the model, which was alone 
impelled by the preflure of the air rifing vertically beneath 
the inclined plane: but, to prove the faft dill more de¬ 
cidedly, I took care that the end of the tube thould be 
placed a coniiderable diftance beneath the bottom of the 
model, and that the head of the model fliould be towards 
me; and, the tube from this pofition being in a contrary 
direction to the track of the model, it was impoflible any 
horizontal impulfe of the blalt (if fuch exilted) could have 
any influence in impelling the model. 
“ This experiment is conclufive as to the faft, that a 
veflel may be impelled by forcing a current of air be¬ 
neath it, fo as to aft in afeending on an inclined plane, 
forming part of its bottom and Item. We have now to 
inveftigate the power of the air fo applied in impelling 
the veflel, the quantity of the air which a fleam-engine 
of known dimenfions can force beneath it to a gtven 
depth, and the beft method of applying the current of 
air fo produced. 
“ One of the beft accounts that we have of the efteft of 
an engine for forcing air is that given by Mr. Roebuck, 
in the Tranluftions of the Edinburgh Royal Society, in 
his account of the blaft-furnaces and air-vault of the 
Devon iron-w'crks. In this account we find that a fteam- 
enginc on Newcomen’s conflruftion, with a main cylin¬ 
der of 48-^ inches diameter, was fuftkient to work a blow¬ 
ing cylinder of 78 inches diameter and feven feet long, 
fo as to produce, when worked to the greateft fpeed that 
the author mentions, nineteen ftrokes in a minute, each 
five feet two inches long. Mr. Roebuck computes that 
a firoke of four feet eight inches expelled 155 cubic feet 
of air under a preflure which fuftained between five and 
fix inches of mercury in the gauge ; confequently a ftroke 
of five feet two inches will produce 171^. cubic feet, 
which, multiplied by nineteen, will give 3257 cubic feet 
for the produce of the engine every minute. Newcomen’s 
engine only afted on the machinery in the defeent of the 
piflon, and then with but a preflure of 7lbs. on thefquare 
inch of the piflon; w hereas one of Watt’s engines afts 
with a force of iolbs. in the round inch of the piflon, and 
with equal power both in its afeent and defeent. Mr. 
Roebuck reckons that .the power of Newcomen’s engine 
above mentioned, was equal to 13,o6albs. each ftroke; and 
it follows, from the above fafts, that an engine on Watt’s 
conflruftion, 26 inches diameter in its piflon, will pro¬ 
duce the fame quantity of air from a blowing cylinder of 
the dimenfions above ftated. 
“ A fleam-engine of Watt’s conflruftion, of 26 inches 
diameter in the piflon, may be carried in even a fmall 
veflel without inconvenience; and, as every cubic foot 
of air produced as mentioned by this engine, will occupy, 
near the furface of the water, when forced beneath if, 
nearly the fame fpace as a cubic foot of water, it follows 
that its re-aftion will be equal to the weight of the water 
it difplaces ; now, as a cubic foot of fea-water is eftimated 
to weigh 64lbs. it follows, that the re-aftion of 3257 cubic 
feet of air, forced beneath the water in a minute, will be 
equal to 2o8,448lbs. and that the primary force of the air 
produced, as deferibed by fuch an engine, will aftt on the 
inclined plane with a power equal to that weight in im¬ 
pelling the veflel. The force of the preflure forward will 
be diminifhed as the angle of the inclined plane on which 
the air a6ts with the horizon is lefs; but then, as the plane 
inuft defcend to the fame depth from the furface, let its 
inclination be what it may, it mult of courfe be longer as 
its angle of inclination is lefs ; and, as the quantity of air 
prefling on it at one time will be greater in proportion to 
its length, what is gained in this latter way will nearly 
compenfate for what is loft in the former. The efteft of 
the preflure of a given quantity of air upwards, againfl 
the inclined plane, will fuffer a reduftion, from its occu¬ 
pying a lefs fpace as it is forced down deeper into the 
water. If it were forced down thirty feet, it would oc¬ 
cupy but about half the fpace it did at the furface, 
and have but half the preflure upwards there; and the 
whole force of its afeent from this depth would be a fourth 
lets on this account than it would otherwife be. In rifing 
from a depth of fifteen feet, one-eighth of its force would 
be loft; and from a depth of feven feet and a half, a fix¬ 
teenth of its force. But the greateft reduftion of the 
preflure of the air upwards will arife from the fmallnefs 
of the refiftance of the water to its motion in a contrary 
direftion to that in which the vefl'el moves: this refiftance 
may, however, be increafed by ufing an inclined plane with 
a fmaller degree of inclination to the furface, for thus the 
air will have to move a greater fpace backwards in afeend¬ 
ing to a given height with a given force. The degree of 
this refiftance of the water to the paflage of the air may 
be computed in fome meafure from an experiment of Dr. 
Hook, by which he found that a ball of wood nfeended 
through fourteen fathoms of w-ater in feventeen f'econds, 
having been firft made to defcend to the fame depth, by 
a ball of lead of the fame weight, in feventeen leconds 
alfo, which is a little lefs than five feet in a fecond, or 
59*- inches. Now, affuming five feet in a fecond to be 
the velocity with which air will move through water, if 
the inclined plane, along which it will have to move in 
afeending, is three times the length of its vertical height, 
the air will then have to pafs through three times the 
quantity of water, in palling backwards in a contrary di¬ 
reftion to the motion ot the fhip, that it would have to 
pals in afeending vertically, and will therefore experience 
thrice the refiftance fora third of the given force ; where¬ 
fore. 
