428 On the Chemical Phenomena of Heat. (June, 
was first raised, DG would be < ED... the particle must move 
on towards A, till ED = DG, then its expansion is a maximum. 
Since in every mass of matter the particles must be .arranged in 
rows running in all possible directions, the whole volume must ex~ 
and. Did the rows all extend in one direction only, the expansion 
could take place in the direction _|_ to it. 
We may now see clearly whence arises the difference in the capa- 
cities of solids for heat. The first cause is the different attractions 
of solid particles for heat; a particle which attracts heat more 
powerfully than another will, ccet. par. be surrounded with a more 
dense calorific atmosphere, and will consequently require a larger 
quantity of the matter of heat to raise it from any one to another 
given temperature; .*. if the interstices in solids were =, there 
would be a difference of capacity ; the second is the distance of the 
particles from each other, on which the altitude of this atmosphere 
depends ; the next cause is the situation of the particles with respect 
to each other; as the temperature of a solid is elevated, the inter- 
stices become enlarged .*. the calorific atmosphere becomes more 
and more extensive, and when the centrifugal force > centripetal, 
a separation takes place ; consequently that part of the atmosphere 
which is principally enlarged is on and near the surface of the par- 
ticle where it is densest .*. during separation a large quantity of 
heat must disappear ; hence all the capacity of every solid increases 
as the temperature is increased. 
From what has been stated respecting solids, it is evident that the 
capacities of = weights of different substances do not represent the 
real capacities ; for as the ratio of the capacities is that of one atom 
of each substance, the capacities ought to be calculated for weights 
in the ratio of the weights of the ultimate atoms: this will appear 
more clearly when the capacities of fluids, gases, and compound 
substances, are examined. 
When the temperature of a solid is elevated to a certain degree, 
DB becomes | AC (Fig. 5), for DF = DH .. the solid just 
begins to lose its cohesion. ‘The most minute addition to the sen- 
sible heat causes an entire separation of the particles, as A BC 
(Fig. 6); and the particles, being now capable of motion in any 
direction, must constitute a fluid. 
Their surfaces now not being in contact, the quantity of caloric 
in each atmosphere has to be increased in that part which is most 
dense ; hence during the conversion of a solid into a fluid, much 
caloric disappears. Were it not for atmospheric pressure, bodies 
would, evidently, only exist in the state of solids and permanently 
elastic fluids; for when the particles are separated, they attract 
caloric more powerfully than they do each other .*. above that tem- 
perature which fuses a solid, its particles, by attracting caloric, and 
continually increasing the extent of their calorific atmospheres must. 
separate ad inf. without becoming previously fluid. This is proved 
to be correct by many experiments. Sal-ammoniac, oxide of 
arsenic, and many other substances, at a certain temperature, vola- 
