CAPTAIN W. DB W. ABNEY ON THE SOLAR SPECTRUM. 
461 
variety of other metals for radiation, extending from X 6600 to 27,000, which we hope 
to submit shortly to the Royal Society. 
Measurement of the negatives. 
The negatives, some 30 in number, were all measured by a micrometer by A. Hilger, 
which was placed at my disposal by the Department of Science and Art. Of the per¬ 
formance of this instrument I cannot speak too highly. It is capable of measuring to 
the tooVoo °f an inch, but for any photographs which I have obtained, or which I 
believe can be obtained, this is too delicate ; the toWoo °f an inch i s the unit which 
alone is necessary, any readings for a well defined line rarely differing more than 1 or 2 
of these units from the mean of a series. In most of the photographs measured, a 
wave-length on Angstrom’s scale is nearly of an inch; any error in wave-length 
through bad measurement would therefore only be effected in the second place of 
decimals : a degree of accuracy which is, for the present, sufficient, since the wave¬ 
lengths are only carried to the first place. 
o 
In obtaining fiducial wave-lengths the numbers given by Angstrom for the visible 
spectra of the higher order of spectrum with which the infra red had to be compared 
have been generally adopted, though in some cases the corrections found in Vogel’s 
map have been utilised when it was found that the interpolation formula better 
agreed with the latter than with the former. Cornu’s map was used for the ultra 
violet comparisons, together with photographs on the normal scale made by Professor 
Rowland. In many instances the wave-lengths have also been checked by photo¬ 
graphs taken on the normal scale, the measures obtained being proportional to the 
differences in wave-lengths. The labour of mapping the spectrum has been much 
augmented by this process, as every difference in the first decimal jalace between 
o 
Angstrom’s and my own determination had to be discussed. 
In the tables the first column shows the number assigned to each line, the second 
column the measures of the photographs, the third the deduced wave-lengths, and 
the fourth check measures taken from other photographs. 
Atmospheric Lines. 
In previous papers by Colonel Festing and myself we have discussed the effect of 
water in the atmosphere, and shown that it has a marked effect in shortening the 
solar spectrum when present in any quantity. We have also shown that the A line is 
not due to aqueous vapour. It has been asserted by Eggeroff that this line is due to 
oxygen. This may be so ; but if it be, the absorption by this gas is very intense, since 
I have found that it does not vary in intensity when observed and photographed at a 
height of about 9000 feet, at Zermatt at the end of June, from that observed in England 
at sea level. That many lines are due to aqueous vapour there is no doubt, but there 
are an equal number of lines which cannot be so disposed of. Part of the group, for 
instance, extending from A to 8200, alter in appearance at sunset and midday, whilst 
