22 ON THE ACTION OF TRANSPARENT BODIES 
peated trials, and as it would have been impossible to replace 
them from the same source, the prosecution of the experi- 
ments with fresh oils, sold under the same name, might have 
introduced a new degree of uncertainty among the results. 
For all the purposes of the Practical Optician, the relative 
position of the substances, where it has been determined by 
experiment, is sufficiently accurate. Numerous combinations 
for correcting the secondary spectrum, may be formed, ei- 
ther by taking two media that have nearly the same ac- 
tion upon green light, while they differ in their powers 
of refraction and dispersion, or by adopting the ingenious 
method discovered by Dr Brarr, and fully described in the 
Transactions of this Society. In the production of perfect 
achromatic instruments, the optician cannot expect much more 
aid from the principles of optics. The great, and almost the 
only desideratum, is to obtain two kinds of glass, which shall 
have the same action upon green light, while they differ suffi- 
ciently in refractive and dispersive power; and it is chiefly 
from the labours of the Chemist, guided by the preceding in- 
quiries, that such a discovery can be expected. 
There is still, however, another source of error, of which 
neither the Theoretical nor the Practical Optician has hitherto 
been aware. It arises from a crystallisation in the glass, which 
is always accompanied with double refraction, and with a va- 
riation of density. This crystallisation, which most frequent- 
ly affects the flint-glass, and which can easily be detected by its 
action upon polarised light, should be carefully removed from 
every piece of glass used in the construction of optical instru- 
ments, by annealing it in an oven of a high temperature, 
where the heat is regularly and very slowly reduced. The ex- 
periments upon which this opinion is founded, will form the 
subject of another paper. 
TABLE 
