New Form of Mountain or other Baron 
end which we had in view, in various plans which were con- 
trived and tried, was to provide some means of following the 
expansions and ‘odntiactighbaal the mercury, arising from chan- 
ges of temperature; so as to furnish the air no chance of getting 
in. his the elastic plate does very well; indeed I may say 
completely: the heat of the hand being enough, in a few min- 
utes, to cause a convex appearance, and the pressure of the at- 
mosphere, on its sana to a cold acm sufficient to make it 
concave. 
‘It was of course: iphorieniolt to. wore: hin solic ond ne conse- 
quence the disk, of as small a diameter as would be compatible 
with the other conditions requisite. After some experiments, 
therefore, to ascertain what elasticity was attainable by hammer- 
ing out the Russia-iron, consistent with uniformity and strength, 
I calculated the following table, shewing the relations existing 
between the clear inside diameter of the disk and that of the 
tube, on the assumption that the range of temperature does not 
exceed 90° Eanes, meer aint ap oe anoceuty:s is agp e% pressed upon 
a the ~~ ‘ fe Seg eS a Pep 
" Ineibe tube-diameter. igs rc ; : ‘or eats rea 
ar 0.90 
ei pL ORG, nals ; 
0.30 136. 
‘ 0.40 be hdeBk r 
( 0.60. ic Geb mg ee TA git « 
ADs iy ois LB: a 
0.75... sibhey! Oe 
These aptvtioias paver bean found to'suit very well in prac- 
tice in the instances where they have been applied. 
‘Thave before spoken of the advantage in precision, which the 
symmetrical disposition: of the float is presumed to afford to the 
measurements. I give here the formula, for calculating the 
amount of correction which is to be applied to make up for the 
partial sinking of the float, ‘and the er elevation of the 
mercury in the tube and cistern. 
It is manifest, that the heavier shee float, vith the same diame- 
ter, the deeper it (and of course the zero of the graduation) will 
sink in the cistern; and also, that. with. a float of constant weight 
Vol, xiv, No. 2.—July-Sept. 1843. 
