326 
PROFESSOR TYNDALL ON THE ACTION OF FREE MOLECULES ON 
The agreement here is close enough to illustrate the law: the greatest difference 
being under 1 per cent. 
At this point the following entry appears in my note book :—“It might have been 
wise on my part to rest content with the comparison of the long and short tubes with 
the lime light as source of heat. But for the sake of completeness I wished to intro¬ 
duce the incandescent spiral. The fulfilment of this wish—that is to say, the 
successful performance of a single experiment—has cost me more than a week’s work. 
Once however committed to the task, I could not leave it incomplete. 
“ The discrepancy between the two tubes was in no case great, hardly ever exceed¬ 
ing 2 per cent. But the difference was uniformly in favour of the long tube and small 
density. Diaphragms were employed, the position of the short tube was shifted, and 
it was finally placed so that the pile should occupy the same position in relation to its 
adjacent end as it did in relation to the adjacent end of the long tube. The dis¬ 
crepancies then disappeared, the absorptions in the two tubes proving practically 
identical. 
“ Many leaves of paper were covered with observations during the week, but it is 
useless to take up time and space in copying them here. One representative observa¬ 
tion will suffice.” 
Sulphuric ether. 
Source of heat—bright red spiral with rocksalt lens. 
Long experimental tube. 
Pressure. Deflection. Absorption. 
1 inch .... 20*0° 23*5 
2 inches . . . 27‘3° 32’1 
Total heat 51‘0° 
Short experimental tube. 
Pressure. Deflection. 
3‘5 inches . . . 17'8° 
7*0 „ . . . 24-8° 
Total heat 49’0° 
Absorption. 
23'4 
32‘6 
The agreement between the two tubes is as perfect as could be desired. 
It is easy to record these experiments, but it is not so easy to make them. On every 
portion of the apparatus—the source, the tube, the thermopile, and the galvanometer 
—extraordinary care must be bestowed to make the experiments strictly comparable. 
The results were checked by taking the total heat after every experiment—a pre¬ 
caution which insured the detection of any variation on the part of the source. When 
the platinum spiral was used, the battery of ten Grove’s cells, employed to render 
