The Total Solar Eclipse of 1900. 83 
at the extreme violet end. In this we seem to have the true flash; but the 
number of lines is very much inferior to what we should in all probability have 
obtained with a better focus. 
A narrow line parallel to the length of the spectrum appears in the ultra- 
violet part of each spectrum. This is due to the planet Mercury, which was 
situated about 2° preceding, and a little south of, the Sun’s centre. 
In order to indicate graphically the successive changes taking place in these 
two series of spectra, a series of straight-line spectra have been derived from the 
prismatic camera negatives, and placed directly under one another, so that the 
gradual fading out of the lines may be seen at a glance. 
To obtain these photographs a very narrow strip through the middle of each 
spectrum was taken (the rest being masked out), and projected in an ordinary 
enlarging apparatus upon a plate, which was moved to-and-fro by clockwork, in a 
direction at right angles to the length of the spectrum. 
The spectra, thus broadened to a quarter of an inch in width, were photo- 
graphed on three plates: Nos. 1 to 6 of the first series being taken on one plate, 
and Nos. 7 to 12 on another, while the third plate carried the five spectra of the 
second series. In fig. 1, Pl. vmt., the whole twelve of the first series are repro- 
duced together, while the five of the second series are shown in fig. 2 of the same 
plate. 
Unfortunately the motion of this clockwork was not quite uniform throughout 
its whole range, which has had the effect of producing a band of greater intensity 
throughout the length of the spectrum. This does not, however, affect the utility 
of the photographs. A more serious defect is the appearance of a number of very 
fine lines across the spectrum, which might, at first sight, be taken for true spec- 
tral lines, but which are really due to the coarseness of the film, each inequality 
being drawn out by the motion of the plate into a fine line. These mock lines 
can, however, be easily distinguished from the true spectral markings, partly by 
their fineness, and partly by the fact that the true lines are for the most part 
traceable through several consecutive spectra. In three cases, viz., in spectra 
2, 8, and 10, a thin dark line may be noticed. ‘These are due to minute pinholes 
in the negatives. 
In this part of the work I have to acknowledge the very great help I have 
received from Mr. E. E. M‘Clellan, third assistant at the Radcliffe Observatory, 
who not only developed the original negatives, but devoted a great deal of time 
and skill to the production of figs. 1 and 2, Pl. vm. 
The measurement of the photographs was carried out at the Radcliffe 
Observatory, using a microscope-micrometer, constructed in 1896 by Hilger. 
In this connexion I desire to express my acknowledgments to Mr. John Evershed, 
F.R.A.S., who was kind enough to supply me with a MS. list of chromospheric 
P 2 
