MR. W. CROOKES OK REPULSION RESULTING FROM RADIATION. 
249 
index ray of light was observed at intervals varying from a few hours to many weeks, 
several times with each powder, and the means were reduced to the standard 
deflection of the lampblacked disk =100. The column headed “water screen inter- 
posed” is also reduced to the deflection of lampblack=100. In comparing the 
two columns, it must be remembered that the actual amount of repulsion on the 
standard lampblacked disk when the water screen is interposed is only y^th of the 
amount obtained when no screen is in the way, the distance of candle and other 
things being equal. In order, therefore, to compare one column with another the 
results in the column headed “water screen interposed” must be divided by 12. 
Table I. — Black Powders. 
No screen. 
Water screen 
interposed (5 millims.). 
Lampblack ( standard disk ) .... 
. . . ioo-o 
ioo-o 
Iron reduced by hydrogen .... 
. . . 101-1 
107-2 
Tungsten reduced by hydrogen 
. . . 96-5 
95-4 
Palladium iodide 
. . . 87-3 
..* 
Mercury sulphide 
. . . 84-0 
94-0 
225. Table II. 
— White Powders. 
No screen. 
Water screen 
interposed (5 millims.). 
Lampblack ( standard disk) .... 
o 
. . . ioo-o 
ioo-o 
Hydrated zinc oxide 
. . . 40-5 
14-0 
Barium sulphate 
. . . 37-4 
4-2 
Magnesia, ignited 
. . . 37-2 
8-4 
Oxamide . . 
. . 36-0 
10-7 
Silica, precipitated 
, . . 33-6 
4-1 
Lead carbonate (white lead) . . . 
. . . 32-5 
9-8 
Mercury sulphethylate 
. . 28-5 
15-1 
Calcium carbonate 
. . ' 28-5 
3-9 
Lead sulphate 
27-6 
4-7 
The powerful absorption for the invisible heat rays which these white powders 
exercise is somewhat remarkable. Assuming that the ultra-red rays from a candle 
are almost entirely cut off by a water screen, the comparatively strong action shown 
in the first column must be mainly due to absorption of the invisible rays of heat, 
and when these are filtered off through water the action is diminished 48 times. t 
A similar examination of Table I. shows that a water screen only diminishes the 
action about 1 1 times. 
In paragraph 201 of a former paper on this subject I gave the result of an 
When dots occur in a column, they mean that no experiment was tried. 
t Omitting lampblack, the average of the first column is 33"5, and that of the second 8'3. But, as 
already stated, the action behind water must be divided by 12, which reduces it to 07, about A-th part of 83*5. 
MDCCCLXXVIII. 2 K 
