358 
Steam Engine. 
blows of a hammer, which caused the whole apparatus to 
tremble, but ’which gradually subsided as the liquid be- 
came warm* Count Rumford supposes, that the benefit 
cial action of the steam depends for the most part on the 
motion described, causeel by it, and therefore proposes 
dividing the vessel into two parts by an horizontal parti- 
tion of thin copper, and causing a slow cmTcnt of cold 
water to pass through the lower division, and to let the 
steam into this lower part, w^hen the upper became too 
hot to admit of a continuation of the strokes from the con- 
densation of the steam ; by which means the same motion 
being continued in the cold water, would be communica- 
ted to the hot liquid through the thin partition* 
The soap made by the operation of the steam, required 
only six hours boiling, whereas sixty hours and more are 
necessary in the ordinary method of making soap* 2 
Athenceum 66* 
Familkir account of the method of estimating the value of 
a Steam Engine in horse -powers as they are called. 
By a correspondent. 
To Mr. NICHOLSON* 
Sir, 
AS your excellent journal is the repository for useful 
information, whether scientific or practical, I have thought 
I should oblige many manufacturers and others of your 
readers, by sending you a very clear report about horse- 
powers, which a friend of mine has communicated to me^ 
and was received by him from an eminent character in 
answer to an enquiry professionally made* 
It clearly appears from this paper, that the calculation 
by horse^powers must be fallacious, unless engineers 
could agree as to the quantity of work they would arbi- 
trarily, in the first instance, ascribe to one horse ; and 
then the expression would be nugatory. And not only 
So, but it would not then be true that the value of a steam 
