SIR W. CROOKES ON THE PHOTOGRAPHIC SPECTRA OF METEORITES. 
419 
exceedingly numerous all along the spectrum. Moreover, the softness of silver 
makes it easy to get strongly coherent poles from its mixture with the aerolite. 
The silver powder I used was specially prepared by a method given to me by 
Dr. A. Scott, F.Tt.S. The purest commercial silver was dissolved in nitric acid, the 
solution evaporated to dryness in a silica crucible, and the solid residue fused till any 
copper nitrate which might be present was decomposed. The mass was extracted 
with water, and the filtered solution of silver nitrate precipitated by means of a 
boiling solution of ammonium formate and acetate into which the silver solution was 
slowly run. After the liquid had become brilliantly clear and the silver well 
agglomerated, the metal was thoroughly washed and dried. The aerolite in fine 
powder was mixed with an equal weight of the silver, and the mixture slightly 
damped and compressed into a block in a hydraulic press at a pressure of 250 
atmospheres. The resulting cake—after drying—was raised to dull redness, and was 
then cut across to form two poles between which to form the electric arc. The poles 
were mounted in special clips in a mechanical arrangement to enable the arc to be 
accurately focussed on the slit and kept, there during the exposure. 
The Arc Spectrum of an Aerolite. 
In order to identify the different elements present I have worked by a method 
unlike that adopted by most observers. Instead of measuring the wave-lengths of 
the lines and establishing their identity by reference to published tables of wave¬ 
lengths, I have made it a rule to obtain, first, a qualitative idea of the constituents 
present, then to photograph the spectrum of the aerolite and silver mixture in juxta¬ 
position with that of each suspected element by a device already described. In this 
way I have been able with certainty to identify every line in the spectrum of each 
aerolite examined. This removes the uncertainty due to irregularities in the 
determination of wave-lengths by different observers. I can take no responsibility 
for the accuracy of the figures by which I identify these lines ; they are copied from 
the most recent and trustworthy sources—had it been possible to reproduce the 
actual photographed spectra that 1 am now exhibiting there would have been no need 
to have printed any figures at all. The spectrum of the aerolite is easily distinguished 
from the mixed silver and aerolite spectrum by reference to a photograph of the 
silver taken separately. 
The advantage of my method becomes apparent if—for illustration—I take one 
element—nickel—common to all meteorites. In the range of my instrument there 
are no less than three hundred and forty strong lines which I have identified, and on 
the aerolite spectrum I have marked all the nickel lines that can be seen, together 
with the wave-length given in the most recent determination—but the point is that 
any line marked is actually present as found by reference to the spectrum of pure 
nickel, and does not rest upon a coincidence of figures. 
VOL. CCXVIT.—A, 3 P 
