RADIANT HEAT, AND ITS CONVERSION THEREBY INTO SOUND. 305 
of 2*44 per cent, was found for air, and an absorption of 3'75 per cent, for hydrogen. 
In the unblackened tube the absorption by air was 14'75 per cent., and by hydrogen 
16 - 27 joer cent, of the total radiation. 
I went over this ground with the utmost care, using invisible as well as visible heat. 
But, substituting plates of pellucid rocksalt for the plates of glass, I failed to realise 
the effect obtained by Magnus. He ascribed the difference between the results 
obtained with his blackened and his unblackened tube to a change of quality in the 
heat, produced by reflection from the interior surface of the latter. With plates of 
rocksalt, however, though the reflection abides, the change of quality does not occur. 
My position, therefore, in regard to these experiments, is similar to what it was in 
regard to those of Dr. Franz. The results obtained with air, oxygen, and hydrogen, 
were, I hold, due to the chilling of the heated glass ends of the tube by the cold gases, 
and the consequent lowering of the secondary radiation. 
It was shown by Magnus himself, and is moreover obvious at first sight, that the 
unblackened tube sent a far greater amount of heat to the glass plate adjacent to the 
thermopile than the blackened one. That plate being more heated by the source, was 
more chilled by the air when it entered. The greater cooling power of hydrogen 
accounts, moreover, for the advance of the supposed absorption from 14*75 to 16*27 per 
cent. With carbonic acid Magnus detected a difference which had escaped Dr. Franz. 
Instead of making the action of this gas equal to that of air, he found in the blackened 
tube an absorption of 8*1.9 per cent., and in the unblackened tube an absorption of 
21*92 per cent, exerted by carbonic acid. Here true absorption mixed itself with the 
effect of mere chilling, while, with still more powerful gases, the effect of chilling 
retreated by comparison more and more. 
Such were the experiments which determined, in the first instance, the attitude of 
this distinguished man towards that portion of my work which related to the action of 
the air and vapour of our atmosphere on radiant heat. In the defence of his position 
he brought to bear all the resources of consummate skill and large experience. His 
position, however, was by no means a wholly defensive one. He dwelt repeatedly and 
emphatically on the dangers—and they are real—to which the method pursued by me 
was exposed. I had closed my experimental tube with plates of transparent rocksalt 
and he urged against me the hygroscopic character of this substance. Placing rock- 
salt beside a vessel of water under a glass shade, he found that it could be rendered 
dripping wet.* Hence his argument, that, instead of measuring the action of vapour, 
I had really measured the action of brine. This, however, I could not admit. I was 
aware of the danger and had avoided it. In many hundred instances the rocksalt 
plates had been detached from my experimental tube while filled with the very 
^ It has been shown by Professor Dewar that the exposure of a dry plate of rocksalt for 5 minutes 
to saturated air sensibly augments the weight of the salt as determined by a delicate balance, 
MDCCCLXXXII. 2 R 
