656 
ME. J. NORMAN LOCKYER ON SPECTRUM-ANALYSIS 
It follows, however, from what has gone before, that the number of lines in a spectrum 
increases as the molecular distance is diminished within the range of atmospheric pres- 
sure ; and there is no doubt whatever that this will go on with greater pressures until a 
continuous spectrum is reached, as in Eeankland’s classical experiment with hydrogen, 
in which the lines broadened — a peculiarity confined, so far as I know, to that spectrum. 
If this be so, with increased pressure we shall get reduced percentages ; and some day 
we possibly shall be able to largely increase our present knowledge by observing the 
pressure at which, with the increased number of spectral lines in the pure vapours known 
to be in the sun, the percentages of the known solar elements, determined by their 
reversed lines, will together make up 100 or thereabouts. It is clear, however, already 
that a pressure of several earth atmospheres would be required. 
Such an inquiry might not so much make us familiar with the existence of new ele- 
ments in the reversing layer as it would enable us to allocate the unnamed lines of the 
spectrum to those already known to exist there ; and it may also be remarked that now 
the absence of any element from the sun is much more easily determined than before, 
as its longest line only need be considered. If that is not reversed the inquiry need be 
carried no further, for all other lines will certainly be absent. I do not wish to imply 
that a trace of an element can be detected in this way, my experiments rather tend in 
the other direction ; besides which it must not be forgotten that we have bright lines 
in the solar spectrum itself, and that in the chromosphere I have recorded bright lines ; 
and Professor Young has recently recorded more in some observations, to which I shall 
take a future opportunity of referring. 
It is a pleasing duty to me to again take an opportunity of expressing my obligations 
to Mr. Friswell for the valuable assistance he has afforded me in these inquiries. All 
the maps have been constructed by him. 
I have also to express my thanks to G. Gore, Esq., of Birmingham, who supplied me 
with the fluorides used in the experiments, and to J. Mellor, Esq., of the Magnesium 
Metal Works, Salford, for a present of a large quantity of magnesium, potassium, and 
other metals, which have also been used in these inquiries ; they were sent anonymously, 
but I have learnt the donor’s name from Professor Poscoe. 
