418 
SIR W. CROOKES ON THE PHOTOGRAPHIC SPECTRA OF METEORITES. 
of occluded gases was much less, but in all cases the calcium absorption was done at 
least twice, so that a very minute percentage of the inert gases would have been 
rendered visible. 
When all the aerolites had been examined a final experiment was made to prove 
the efficiency of the method. A little samarskite was powdered and treated in the 
same way. On absorbing the liberated gases by heating the calcium—as soon as the 
pressure had been reduced to a few millimetres—the complete spectrum of helium 
was revealed. 
The only gases recognised were the compounds ol carbon hydrogen, and in one or 
two cases only H 2 S ; this latter is somewhat remarkable, as at the time its spectrum 
appeared to lie novel ; the other aerolites were most carefully examined for it without 
success. 
It was found impossible to get an induction spark to pass betiveen two fragments 
of an aerolite, so attempts were made to make use of an arc. A fragment of the 
stone was ignited in a cup hollowed out of the lower carbon 
pole—the upper electrode being a carbon rod. As shown 
in fig. 5, if an arc was struck between the two carbons by 
momentary contact with a third rod of carbon, the fragment 
of stone quickly became hot and ultimately fused and formed 
a molten sphere ; the arc could then be maintained between 
this sphere and the upper carbon. Photographs were 
obtained of the resulting arc and the spectra of several 
of the meteoric stones examined. The plan followed was to 
arrange three stops or apertures that could he adjusted in 
front of the slit, and by their aid to project upon the photo¬ 
graphic film, first the spectrum given by iron; then the 
spectrum of the aerolite ignited between two poles of carbon ; 
and below that the spectrum given by carbon poles alone. In 
this way it was easy to observe any lines in the aerolite 
spectrum that did not appear in either the iron or the carbon, and thus the 
composition could be determined. This method was not altogether satisfactory 
owing to impurities in the carbon rods and the consequent uncertainty of some of 
the observed lines being due to constituents of the aerolite or to impurities in the 
carbon. I therefore tried other means. 
The aerolite was finely powdered and mixed with a metallic powder, the composition 
of which was known. The mixture was strongly compressed in a hydraulic press, 
and from the solid mass thus obtained a pair of poles were fashioned. 
Iron was one of the first metals used for experiment, but was discarded owing to 
the fact that iron is a plentiful constituent of stony meteorites. Silver was finally 
chosen because of the comparative ease of getting it pure—also because the lines in 
any part of its arc spectrum are few in comparison with the lines of iron which are 
