248 ON THE LIGHT OF THE MOON AND OF THE PLANET JUPITER. 
The following diameters of Bs, measured in various directions, show the general char- 
acter of the irregularities of figure to which these glass globes are liable. 
in. in. 
Diameter of B, = 3.501 — 2 x 0.034 
8.506 « 
3.500 ue 
3.503 = 
3.510 " 
Small irregularities of form and in the quality of the reflecting surface are elimi- 
nated by reversing the bulb, so as to present the same part alternately to each object. 
'The absorption being equal for both images, its effect will be eliminated in the ratio of 
the two lights compared. : | 
The convenience of this photometer is its principal recommendation, especially when 
the observations are made at night and in the open air, under various annoyances 
which would prevent the use of any but a very simple apparatus. A polished sphere 
has the property of reflecting an equally bright image of a distant object in all direc- 
tions; but since the proportion of light lost by absorption may be liable to vary some- 
what at different incidences, it is better to place the eye in such a position for viewing 
the two images, that a line drawn from it to the centre of the sphere may make equal 
angles with the directions of the two objects from the same centre, so that the angles 
of incidence for the light reaching the eye will thus be the same for both. 
Care has been taken to place the bulb in the axis of the pencil proceeding from the 
lamp aperture, so that the light compared should come from the centre of the flame, 
and pass the aperture at nearly right angles to its plane. 
If dl be the quantity of light incident from the lamp disc upon an element of sur- 
face d e, placed near the axis of the pencil, and perpendicular to the line d, joining de 
= n. centre of a disc, d? will remain sensibly constant for small changes in the 
direction of d relatively to the axis. The whole quantity of light, incident upon a 
sphere having a radius £, small compared with d, will then be obtained, as in (10). 
1 
D 
2 
Sw 
"iade d is = riens quantity of light which would emanate from the disc, if it shone 
m all iisen with the same brightness that it does upon de when placed in the axis 
of the pencil. So long as the comparisons are confined to the light incident from the 
