ON METEORIC DUST. 91 



precipitated by rain and snow. We are concerned here chiefly with the 

 magnetic particles contained in this dust. These particles are of various 

 shapes, but the most remarkable form is perfectly spherical, which at 

 once conveys the obvious information that the particles at one time must 

 have been in a state of fusion. These spherules have been found in the 

 snows on the slopes of Mont Blanc, at a height of nearly 9,000 feet in the 

 sediment of rain collected at the Observatory of Sainte-Marie-du-Mont, 

 and in the dust collected at different elevated positions. There are other 

 particles, not spherical but of equally characteristic forms, which generally 

 accompany these spherules, and some of these shapes we find in the 

 iron dust extracted by Nordenskjold from the sand of the polar regions. 

 If we confine ourselves, however, at present to the spherical particles, 

 and accept the conclusion that they have been in a state of fusion, we 

 are practically reduced to three alternatives. The particles may be of 

 volcanic origin, they raay have been fused in our terrestrial fires, or they 

 may be meteoric. All the volcanic dust which the writer has had at his 

 disposal was carefully examined under the microscope, and its appearance 

 was found to be altogether diffei'ent from the supposed meteoric dust. 

 Such also seems to be the conclusion arrived at by Tissandier. No iron 

 spherules have, as far as I know, been found in volcanic dust. 



The smoke issuing from the chimneys of our manufacturing towns 

 can and does contain iron particles similar in appearance to those to 

 which Tissandier ascribes a meteoric origin. That some of these particles 

 are found very far from any terrestrial sources which can produce them 

 would not perhaps tell conclusively against their terrestrial origin, but 

 chemical analysis seems to settle the point. The iron particles issuing 

 from our chimneys contain neither nickel nor cobalt ; while these metals 

 were found by Tissandier to exist in the microscopic magnetic particles 

 found in rain-water collected at the Observatory of Sainte-Marie-du-Mont. 

 We are, therefore, driven to ascribe a cosmic origin to these particles. 



2. The author of this Report has, during the last year, made a few 



microscopic investigations of small iron particles found in different places. 



He has obtained in the first place sand col- ^ , „,,-,,. 



1 J. J r J.T. J J. • J.1, • 1,1, 1, 1 Fig. 1- — Globular Pieces of Iron 



lected from the desert m the neighbourhood ^^ geen under the microscope 



of the Great Pyramids. This sand contains (objective J inch), March 16, 



an appreciable quantity of magnetic particles 1882. Specimen of Sand col- 



ai. • T 7^ AAA^ T? • J V J.1 lected near the Pyramids m the 



part m 144,000). Examined by the Desert of Sahara. 



microscope the greater part of these particles 



were found to be angular in shape, and there 



can be no doubt that they form an integral 



part of the sand, and are due to the debris 



of magnetic rocks. But here and there we 



find perfect spheres of iron exactly like those 



described by Tissandier and about the same 



diameter, that is, 0'2 to O'l millimetre; 



some of them are even larger. Fig. 1 shows 



some of these spherules as seen under the microscope. Figs. 2 and 3 



give another form of frequent occurrence. 



The greater part of these latter particles are metallic iron, as is shown 



by the deposition of copper on them when a drop of a solution of sulphate 



of copper is added. In fig. 2 all the light parts represent the metallic 



iron. The total quantity of this metallic iron must, however, have been 



small, for it could not be traced by chemical analysis. 



