588 
MR. J. NORMAN LOCKTER ON THE 
A thin layer of vapour of equal brightness throughout, and resting on the photo¬ 
sphere, will give short arcs having the appearance shown at b in fig. 11 ; as we are 
observing a spherical shell, the appearance would not be different from that at a, but 
the edges of the arc might be expected to be more sharply marked. 
In the case of a concentric shell some distance removed from the photosphere, its 
representative in the spectrum of the cusp will be a relatively long arc as at c in fig. 11, 
brightest near its outer edge, and suddenly dimming for the reason above stated. 
Different layers of this kind will give arcs struck with different external radii. 
The arcs due to all but the brightest layer's, however, will be lost in the general 
illumination of the field, and since the brightest vapours will be near to the photo¬ 
sphere. the differences of radii will not be very great, and as we are dealing with short 
arcs, these differences of radii will not help us to distinguish the outer shells in this 
way. 
This suggests another consideration which may perhaps help us eventually, when 
higher dispersions are employed to distinguish concentric shells, if they exist. As 
the arcs due to them are brightest on the outside, they will appear to occupy places 
in the spectrum which do not correspond to their true wave-lengths. At the same 
time they will be longer, so that it is in the case of long arcs that we might expect 
to find departures from the true wave-lengths; if we found, for instance, three lines 
near iron lines with a constant difference from the true wavedengths of the iron lines, 
we should be justified in regarding such lines as iron lines displaced in the manner 
indicated. 
Comparison of the Methods in the case of Vapours close to the Photosphere. 
The prismatic camera offers special advantages for the investigation of the vapours 
which lie nearest to the photosphere. 
(1.) The spectrum of even a very shallow layer will be represented by arcs of con¬ 
siderable length in a photograph taken with the prismatic camera at the beginning or 
end of totality, the length of the arc corresponding to the whole of the layer 
uncovered by the moon at the time. With a slit spectroscope the lines will be 
relatively much shorter, even assuming that the slit can be placed exactly at a 
tangent to the moon’s limb. This is shown in fig. 12, in which the continuous circle 
represents the moon, and the dotted one the disc of the sun. A layer of vapour 
resting on the photosphere is represented by the outer portion of a dotted circle. In 
the prismatic camera this layer would be represented by an arc of length ab, while, 
with the slit spectroscope, under the most favourable conditions, it will only be 
represented by a line of length cd. 
No special adjustment of the instrument is necessary to enable such photographs 
to be obtained, whereas, with the slit spectroscope, the most accurate adjustment 
would be necessary. 
