5 
optical properties and structure of heated glass , &c. 
and bottle glass ; and that the contraction of the internal parts , 
in consequence of which the vacuities are produced, is not 
necessary to that arrangement of particles upon which these 
properties depend. In the flint glass drops, such as ABC, 
fig. 1, there is sometimes only one vacuity, in the thick part 
at E, and as the slender extremity C is perfectly cold before 
the vacuity E is formed, and when the glass round E is red 
hot, it is obvious that the part C has suffered no contraction, 
and is in the same state as the crown glass drops obtained by 
Dr. Hope. But the extremity 'C has a more perfect struc- 
ture than the bulb AB, as it possesses distinct neutral axes : 
hence we may infer that the crown glass drops, without va- 
cuities, will exhibit neutral axes in every part of their length ; 
that their structure is more uniform than that of flint and 
bottle glass drops ; and that the difference between the spe- 
cific gravity of the drop, and that of the annealed crown glass 
from which it is made, will afford a correct measure of the con- 
traction which the glass experiences, in passing, by a gentle 
gradation of temperature, from the fluid to the solid state.* 
• When a piece of red hot steel is plunged in cold water, it experiences a diminu- 
tion of density analogous to that which takes place in drops of melted crown glass. 
Mr. R. Pennington found that a piece of steel, which, when soft, measured 2.769 
inches had expanded to 2.7785 inches, after being hardened by immersion, when red 
hot, into cold water. Mr. Cavallo gives the following measures, without mention- 
ing by whom the experiment was made. 
Specific gravity of soft steel hammered - - 7.840 
■ ■ ■■ - of soft steel hammered, and hardened in water 7.818. 
In all these cases the particles are held in a state of unnatural constraint by the sud- 
den induration of the external coat j and therefore it is probable, that neither the 
glass nor the steel will expand by any moderate accession of temperature. If this 
conjecture be well founded, it will enable us to supply one of the greatest desiderata 
in the arts, a pendulum of invariable length . 
