594 
DR. J. T. BOTTOMLEY ON THERMAL 
Ill two cases, the highest polished copper and the sooted globe, the surfaces are 
similar to those used by Macfarlane. In the other two cases, “ copper polished 
and finely lacquered” and “copper silvered,” the surfaces are new. It is interesting 
to note that the results for lainp-hlack differ very appreciably from those obtained by 
Macearlane. All my experiments show that there is difference in radiation from 
different kinds of lamp-black ; though I doubt, considering the way in which “ lamp¬ 
black radiation” is often referred to, whether the ideas commonly held quite corre¬ 
spond with this view. 
On the other hand, the column showing loss of heat from the “ copper ball polished 
bright,” agrees, with almost perfect exactness, with the numbers obtained by 
Macfarlane under the same circumstances. This confirms, in a very important 
way, both sets of results. It is quite natural that if both are accurate the two 
should agree perfectly, because to obtain a well-polished surface of copper again and 
again in much the same condition is comparatively easy. It was found very difficult, 
however, to keep the polished surface of copper unoxidized, even for a few hours. 
Hence, I tried a thinly-laccj^uered copper surface. The lacquer, of course, alters the 
condition entirely from that in which the surface is highly polished and unlacquered, 
but the condition was permanent, and, indeed, it is only in this condition that briglit 
copper is ever met with in the arts or in the laboratory. It might be interesting to 
try a dull surface of copper lacquered, just because it is the most common of all. 
The silvered surface was the first I had tried of the kind. It was bright ; but no 
special pains were taken (such as I afterwards took) to get the very highest possible 
polish. It is worth observing that the bright fresh copper shows a slightly lower rate 
of loss of heat, probably lower radiation, than the silver-coated globe, while the increase 
with temperature of the rate of loss is almost the same in these two cases, but very 
different in the other cases. 
In all these cases the air was at the natural pressure of the day. It was not 
artificially moistened, as was the case in Macfarlane’s experiments, nor artificially 
dried, as in the case of my own experiments which followed. 
The experiments of Macfarlane, and those just referred to, were made with the 
copper globes suspended in an enclosure of large dimensious, and at ordinary atmos¬ 
pheric pressure. To make use of a Spreugel vacuum, however, and to compare 
results obtained with vacuum of various degrees of pressure and at full air pressure, 
it is necessary, according to the theory of Crookes’ results, to have an enclosure 
of moderate dimeusions, so that the condition may be reached in which the length of 
free path of the gas molecules becomes comparable with distances, from point to 
point, of portions of the enclosure. 
For the reasons stated in my former paper,* I also considered it essential that the 
enclosure should be metallic. I considered whether a glass globe lined inside with 
* ‘Phil. Trans.,’ 1887, p. 444. Uncertainty, ■svitli a substance so badly conducting and so little 
diatliermanons as glass, as to the temperature of the inner skin of such an enclosure. 
