442 W. A. Norton—Molecular and Cosmical Physics. 
ro 0 
exceptions among solids, and water below the temperature of 
39° Fr. among liquids. For heat contracts india-rubber and 
water (below 39° F.); and the tensile strength of bar-iron aug- 
ments as the temperature rises, until it reaches about 500° F. 
Beyond this a farther rise of temperature has the opposite effect 
From this we may infer that bar iron in cooling, from a high 
temperature, passes back from the secondary into the primary 
condition at about 500°. Water does the same at 39°. The 
temperature at which the change occurs in india-rubber is not 
known. 
10. We have now distinctly recognized the important theo- 
retical principle, that primitive molecules are subject to a change of 
physical condition, with varying circumstances; and that tas 
change consists ina recess of the molecular envelope from the cen- 
tral atom, or an approach toward it, and is attended with a bs 
in the intensities of the several molecular forces exerted by the mole 
cule at a given distance, and in that of the effective molecular achon 
at any supposed distance. This principle furnishes the key 
the explanation of a multitude of general and special phenom: 
ena, which have hitherto remained inexplicable, on the preva 
lent notion that the atoms of bodies have certain inherent, 
unvarying forces—repulsive at one distance and attractive at 
another, or indeed upon any supposition of vibratory or other 
modes of atomic motion as sources of special forces. ; 
It is important to observe here, that this change of the physi: 
cal and mechanical condition of molecules, may, 10 § 
eases, be either transitory or permanent; and that any ¢ a 
effected may either wholly or only partially por off when 
e seen on atten 
tively considering the change in the condition and extent of . 
the portion of the ethereal atmosphere of a molecule lying 
return to its original position, or more probably may stop short 
of it more or less. : 
11. These principles serve to explain the distinction between 
Saty and imperfect elasticity. It has been conclusively a 
ished by the experiments of Hodgkinson, Chevandier 
Wertheim, and others, that many, and probably the great 
majority of solids, when relieved of any strain to which they “a 
ave been exposed, do not entirely recover themselves, first 
ave a certain set; which, from being very minute when it 
