ME. W. CROOKES ON REPULSION RESULTING- FROM RADIATION. 
357 
changes, d is a cardboard tube, blackened inside, for the light to pass along ; it has a 
movable shutter ( e ) at the outer end, which can be opened or closed by a touch of the 
finger without shaking the apparatus. The torsion-balance is firmly fastened to the walls 
of the room, which are very thick and firm, whilst the bottles and rest of the apparatus 
are supported by the floor ; vibration caused by walking about or touching any of the 
other pieces of apparatus is therefore not communicated to the torsion-beam, f is a 
small lamp, a ray from which, passing through a narrow slit, falls on the inclined 
mirror ( g ), whence it is reflected to the suspended mirror of the torsion-apparatus. The 
light is now reflected back again to the mirror (g) and the graduated scale (A), where 
its position indicates the movements of the torsion-beam. The mercury-pump (not 
shown in the figure) is at some distance from the apparatus, so that the radiation from 
my body might not affect the apparatus when the pump was set working. 
The ray of sunlight (?') falls first on the silver mirror (j) of a heliostat moving by 
clockwork ; thence it is reflected to the slit k, along the tube Z, through the prisms m , 
and the lenses n. o is a large screen, with a narrow vertical slit in it, so as to allow only 
the part of the spectrum I wished to experiment with to pass on to the apparatus. 
jp is another screen extending for a considerable distance, and intended to prevent extra- 
neous light from entering the apparatus when the shutter ( e ) is opened. It has a square 
aperture of such a size that a beam of light passing through it will just cover the pith 
surface (13 millims. square). The slit was kept half a millimetre wide during the 
experiments ; the focus of the lenses was so adjusted as to form a sharp image of the 
spectrum lines at the place occupied by the square of pith on the torsion-beam. The 
slit and prisms are on a firm stand, capable of rotating through a small angle on a centre 
near the prisms. By this means any desired portion of the spectrum can be thrown on 
the pith without disturbing the focus. This necessitates an alteration in the adjustment 
of the heliostat, but that is soon effected. Slight alterations in the position of the 
spectrum can be made by moving one or both of the lenses sideways. 
The distance between the slit and prisms is 16 inches ; from the prisms to the pith 
surface is 7 feet 8 inches. The length of the spectrum here is 200 millims. from the 
lines A to G. The apparatus was fitted up in a room which admitted the sun in the 
right direction from about 10 a.m. to half-past 1 p.m. The heliostat was sufficiently 
good to keep a beam of light on the apparatus the whole of that time. 
188. The past summer and autumn have been very unfavourable for spectrum researches 
of this character. My first accurate experiments with the above described apparatus were 
tried on July 26, and between that date and the middle of September the apparatus was 
kept in good adjustment, so that use could be made of the comparatively few occasions 
when the sun was sufficiently clear. I find that, owing probably to a variation in the 
aqueous vapour in the atmosphere or to slight haze, the observations of one day are 
not readily compared with those of another. Isolated experiments on one part of the 
spectrum are therefore of slight value, and my endeavour has been to embrace as wide 
an extent of spectrum each morning as I was able. A cloudless sky is not always the 
MDCCCLXXVI. 3 D 
