CELESTIAL SPECTROSCOPY. 99 
that the image of the réseau will appear together with the images of 
the stars when the plate is developed. This great work will be divided 
according to their latitudes among eighteen observatories provided 
with similar instruments, though not necessarily constructed by the 
same maker. Those in the British dominions and at Tacubaya have 
been constructed by Sir Howard Grubb. 
Besides the plates to form the great chart, a second set of plates for 
a catalogue is to be taken with a shorter exposure, which will give 
stars to the eleventh magnitude only. These plates, by a recent de- 
cision of the permanent committee, are to be pushed on as accurately 
as possible, though as far as may be practicable plates for the chart 
are to be taken concurrently. Photographing the plates for the cata- 
logue is but the first step in this work, and only supplies the data for 
the elaborate measurements which have to be made, which are how- 
ever less laborious than would be required for a similar catalogue with- 
out the aid of photography. 
Already Dr. Gill has nearly brought to conclusion, with the assist- 
ance of Prof. Kapteyn, a preliminary photograhic survey of the south- 
ern heavens. 
With an exposure sufficiently long for the faintest stars to impress 
themselves upon the plate, the accumulating action still goes on for 
the brighter stars, producing a great enlargement of their images from 
optical and photographic causes. The question has occupied the atten- 
tion of many astronomers, whether it is possible to find a law connect- 
ing the diameters of these more or less over-exposed images with the 
relative brightness of the stars themselves. The answer will come out 
undoubted in the affirmative, though at present the empirical formule 
which have been suggested for this purpose differ from each other. 
Capt. Abney proposes to measure the total photographic action, in- 
cluding density as well as size, by the obstruction which the stellar 
image offers to light. 
A further question follows as to the relation which the photographic 
magnitudes of stars bear to those determined by eye. Visual magni- 
tudes are the physiological expression of the eye’s integration of that 
part of the star’s light which extends from the red to the blue. Photo- 
graphic magnitudes represent the plate’s integration of another part of 
the star’s light, namely, from a little below where the power of the eye 
leaves off in the blue to where the light is cut off by the glass, or is 
greatly reduced by want of proper corrections when a refracting tele- 
scope is used. It is obvious that the two records are taken by different 
methods in dissimilar units of different parts of the star’s light. In the 
case of certain colored stars the photographic brightness is very differ- 
ent from the visual brightness; but in all stars, changes, especially of 
a temporary character, may occur in the photographic or the visual 
region, unaccompanied by a similar change in the other part of the 
spectrum, For these reasons it would seem desirable that the two sets 
