394 
MR. J. EVERSHED ON WAVE-LENGTH DETERMINATIONS, ETC., 
intrinsically more luminous and denser, give fainter arcs because of the excessively 
small angular width of the radiatino’ area. 
It is to be borne m mind that the flash-sjDectrum arcs or “lines” obtained with a 
jii'ismatic camera are not true images of the strata producing them, but diffraction 
images more or less enlarged by photographic irradiation. The deej)est lavers 
subtend an angle of only 2", representing on the sun about 900 miles, and mono¬ 
chromatic radiations from such strata will produce lines which are as narrow as 
instruments of ordinary resolving power are capable of defining. The diffraction 
image of a shallower stratum will not be narrower in proportion but simply less 
bright. A layer of gas 9 miles in depth would, if sufficiently dense, be capable of 
producing strong absorption lines in the solar spectrum, yet it would have to be 
100 times more luminous than a stratum extending for '1" above the photosphere 
to give equally strong bright lines in the photograph, even were the moon to remain 
stationary at the contact and allow of equal exposures. 
But it is obvious that, for the shallosv layers, the time during which the plate is 
exposed to any radiation is proportional to the depth of the luminous layer, the 
advancing limb of the moon cutting off the shallow strata during the exposure ; 
consequently, for an exposure timed from the moment of second contact and lasting 
until the whole dej)th of '2" had been occulted, the plate would be exposed to the 
9-mile stratum for only whole exposure. This layer would therefore 
need to be 10,000 times more luminous to give lines of equal intensit}" to those of the 
deepest layers. 
[In the ahsorption spectrum the relative i 2 itensities between the different elements 
must depend on the total number of absorbing molecules of each element encountered 
by a ray of liglit in its passage outward from the photosphere, and not on the 
relative state of diffusion or depth of the different gases. 
To illustrate these points, which appear to me to be of great importance in 
elucidating the relationship between the emission and absorption spectra, we may 
consider the typical cases of iron and of titanium in the chromosphere. 
The total number of iron molecules encountered by a ray of photospheric light 
probably vastly exceeds the number of titanium molecules, since the absorption lines 
of the former are among the strongest and broadest in the spectrum, whilst those of 
the latter are narrow and mostly weak lines. But the iron vapour is concen¬ 
trated in a stratum close down upon the photosphere, whilst titanium is diffused 
through the entire depth of the chromosphere. Hence the apparent intensities of 
the emission lines of iron fall considerably below those of titanium. 
If the relative distril)ution of these two elements were reversed without changing 
the total quantity of eacli, the absorption lines ^A'ould probably not be materially 
altered in intensity, but the bright lines of iron would then rival those of calcium in 
H and K, whilst the titanium lines would be difficult to detect in the flash spectrum.— 
March 10, 1901.] 
