[ 198 ] 
XXIV. Defcription of a new Hygrometer tr 
By Mr, John Smeaton, F. R, S, 
• Re*id March 2 ^T_T I N G foiTie years ago attempted 
1771- -J^ J[ to make an accurate and feniible 
hygrometer, by means of a hempen cord, of a very 
confiderable length ; I quickly found, that, though 
it was more than fufficiently fufceptible of every 
. change in the humidity of the atmofphere, yet the 
cord was, upon the whole, in a continual flate of 
lengthening. Though this change was the greatefl: 
at hrft, yet it did not appear probable that any given 
time would bring it to a certainty j and futhermore 
it feemed, that, as the cord grew more determi- 
nate in mean length, the alteration by certain dif- 
ferences of moifture grew lefs. Now as, on conli- 
dering wood, paper, catgut, &c. there did not ap- 
pear to be a likelihood of finding any fubftance 
futficientlv fenfible of differences of moifture, that 
, would be unalterable under the fame degrees there- 
of ; this led me to confider of a conftrudtion which 
would readily admit of an adjuftment ; fo that, though 
the cord whereby the inftrument is adfuated may 
be variable in itfelf; both as to abfolute length, and 
difference of length under given degrees of moif- 
ture, yet that, on fuppofttion of a material depar- 
ture from its original feale, it might be readily re- 
4 ftorcd ' j 
