ME. W. CEOOKES ON REPULSION RESULTING FROM RADIATION. 345 
("sol. ammonio-sulpliate of^O-O 
20‘0 secs, for 1 revolution. 
1 candle, 4 inches off, behind 
\ copper, 7^ millims. thick. 
154. The instrument is thus seen to be capable of very extended use as a measurer of 
radiation of any desired kind. Unlike the instrument described in paragraph 135, it 
cannot be used for actually balancing one quality of light against another; and the 
method of taking an observation is not so accurate, for it is less easy to count revolu- 
tions per second or per minute than to observe the movement of a spot of light along 
a graduated scale. There are besides many causes which tend to interfere with the accu- 
racy of the indications of this form of instrument. But, notwithstanding these draw- 
backs, I think the radiometer is likely to be a more popular form of light-measurer. 
It requires no adjustment, and is always ready to be observed, whilst there is a peculiar 
charm in using an instrument which is constantly in active work. With the exception 
of the comparison by balancing one light against another, all the observations mentioned 
in paragraph 140 can be taken with the radiometer, and it is besides capable of appli- 
cations of its own. I will mention one, although others easily suggest themselves. 
As the radiometer will revolve behind the orange-coloured glass used by photo- 
graphers for admitting light into their so-called dark room, it is only necessary to have 
one of these instruments in the window to enable the operator to see whether the light 
entering his room is likely to injure the sensitive surfaces there exposed ; thus, having 
ascertained by experience that his plates are fogged or his paper injured when the revo- 
lutions exceed, say, one in three seconds, he will take care to draw down an extra blind 
when the revolutions approach that number. In photographic operations a radiometer 
may be placed in some convenient spot near the object to be copied. Having ascer- 
tained, once for all, how many revolutions the instrument makes whilst a good negative 
is being taken, the operator need in future take no account of the variation of light, but 
simply expose for the same number of revolutions, with a certainty that his negatives 
will all be of the same quality. 
For the more important work of gas-testing probably the bar-instrument already 
described (135) will be more valuable ; although, even for this purpose, the radiometer 
will be found to give very rapid and trustworthy indications. 
155. I have already mentioned that the motion of the radiometer depends on a dif- 
ferential action of radiation on the black and white surfaces. To obtain rotation in the 
ordinary way the black must be repelled with more energy than the white ; and this 
appears to be the case with all the luminous rays. In the case of dark heat, however, 
