36 & 
Steam Engine. 
ptopriety of making a few experiments with a view of 
determining how far a load upon the back of each draught 
horse, would assist him in his labour. Nor can it be re* 
quisite to point out in what way, by means of such steel- 
yards properly applied to waggons, &c. upon tolerably 
smooth roads, and two horses marching abreast, (one 
drawing the load, the other raising the weight,) experi- 
ments might be instituted to ascertain the magnitude of 
the efforts of horses when drawing in rectilinear paths. 
Judicious experiments having these purposes in view, 
would certainly be beneficial, as they would enable us to 
tell what advantages might be expected from the labours 
of this useful quadruped in different circumstances. But 
with respect to the adoption of “ horse -power^^ as a unit 
of force in estimating the power of steam engines, &c. I 
confess that if it were as well known, and as un variable as 
the length of the day of the equator, I should feel an aver* 
sion to applying it to any such purpose. It is a common 
measure arbitrarily adopted, which has no necessary con- 
nection with the subject that is referred to it, which does 
not in any respect facilitate the computation of the powers 
of an engine, and which may, without proper caution, 
lead to considerable errors in the conclusions deduced 
from it. 
Before I close this letter, already perhaps too long, I 
beg permission to say a few words respecting the mea- 
sure which is generally employed to determine the me- 
chanical effect produced. This is the measure of the de- 
servedly celebrated Mr, John Smeaton, who says that 
the weight of a body multiplied by the height through 
which it descends, while driving a machine, is the only 
proper measure of the power expended; and that the 
weight multiplied by the height through which it is uni- 
formly raised is the only proper measure of the effect 
! produced,” Mr, Smeaton was led to the use of this 
