18 REPORT—1840. 
more prominently forward by some remarks in the Physical 
Section of the British Association at Liverpool in 1837*. In my 
former report I had thrown out some suggestions both as to the 
want of such a series of experiments, and as to the fundamental 
difficulty arising from the variety of causes which must influence 
the results; but more especially the differences of thickness 
in the coatings, which in the ordinary mode of operating could 
not be estimated, yet must greatly modify the effects. 
With reference to the necessity of equalizing the coatings, 
Mr. Bache refers to an important observation of Leslie, viz. 
that radiation takes place not merely from the actual surface, 
but from a certain depth, or lamina of the surface, the thickness 
of which is quite appreciable in good radiators, and differs for 
different substances. 
Proceeding upon this fact, the author justly observes, that 
** the radiating powers of substances would not be rightly com- 
pared by equalizing their thicknesses upona given surface, nor 
by equalizing their weight; but by ascertaining for each sub- 
stance that thickness beyond which radiation does not take 
place.” 
It is then on the original application of this fundamental 
idea that his whole series of experiments is conducted. 
Upon this principle the first object was to obtain some data as 
to thicknesses of different pigments necessary to be employed. 
The method adopted throughout was to employ tin cylinders 
of the same size, filled with hot water, and having thermome- 
ters inserted through a hole in the top; while their surfaces 
were coated with the different substances under trial. The ra- 
diation was estimated by the observed rates of cooling. 
To find the critical thickness of the coating just spoken of, 
the time of cooling a certain number of degrees was accurately 
observed, first with a thin coating, then with an additional 
layer of the pigment, and so on, until it was found that addi- 
tional thickness did not increase the rate of radiation, but 
began to diminish it; thus each coating was adjusted precisely 
to that thickness at which it produced its maximum effect. 
Every precaution to ensure accuracy appears to have been most 
diligently taken, and several series of preliminary experiments 
are recorded for the purpose of ascertaining the limits within 
which the precision of the results may be relied on. A standard 
cylinder, coated with aurum musivum (as being found not 
liable to tarnish or alteration), was used inall the experiments, 
and the effect of each coating compared with this under similar 
circumstances. 
* See Report, 1837. Sectional Proccedings, p. 20. 
