252 A. LIVERSIDGE. 



Amongst the magnetic particles found in the dust are 

 some which are perfectly spherical, these have been found 

 in the snow of Mount Blanc at a height of nearly 9,000 ft. 

 and in the dust collected at various other places. The 

 spherical particles have evidently been fused — no similar 

 ones have ever been found in volcanic dust ; the smoke 

 from factory chimneys does contain similar particles, but 

 such contain neither nickel nor cobalt, whereas Tissandier 

 found both these metals in the magnetic particles collected 

 at the Observatory of Saint Marie-du-Mont. " We are 

 therefore driven to assume a cosmic origin to these 

 particles." 



Prof. Schuster examined the sand from the desert near 

 the Great Pyramids and found magnetic particles present 

 in the proportion of 1 in 144,000. Most were angular and 

 evidently derived from magnetic rocks; but with them 

 were also some perfect spheres of iron, similar to those 

 described by Tissandier, and about *2 to *1 millimetre in 

 diameter. Prof. Schuster naturally assumes that these 

 spherical particles have been derived from the fused sur- 

 faces of meteorites in their passage through the atmosphere. 

 He accounts for their nonoxidized condition partly by the 

 fact that nickel iron alloys are less readily oxidized than 

 iron, also that the fusion and separation from the meteorites 

 probably took place in the upper regions of the atmosphere 

 where the amount of oxygen is exceedingly small ; at a 

 height of only 200 kilometres the percentage of oxygen in 

 the air is only about 0'8. He then refers to the presence 

 of an unknown gas in the atmosphere indicated by a green 

 line shown in the Aurora borealis spectrum — he assumes 

 that this gas must have a very low density — supposing that 

 it be as light as hydrogen, and that at the surface of the 

 earth its quantity per cubic centimetre is only the one 

 millionth part of the oxygen present in the same space, it 



