254 A. LlVERSIDGE. 



although imperceptible rain of such meteoric dust gently 

 descending upon the earth's surface, the amount under- 

 going considerable variation, at times being much greater 

 than at others, and from all accounts the fall in the month 

 of November is the most abundant. This meteoric dust is 

 apparently of the same general nature, and is probably 

 derived from the same sources as the larger masses known 

 as meteorites. 



Roof Dust, collected from the beams under the roof of 

 the University, February 24th, 1882. The proportion of 

 magnetic material separated from one lot of this dust was 

 l*3°/o, another lot gave 1*7°/°. Most of it acquired a metallic 

 lustre when ground in the mortar, some of this was fused 

 with caustic soda and the metallic residue was found to 

 contain cobalt and nickel. 



Mud from a Cistern. — Collected March 3rd, 1882, from 

 a rain-water cistern (open) under the roof of the University 

 building. When dried it yielded a grey coloured powder, 

 containing sand grains, vegetable fibres, etc. The magnetic 

 portion was fused with caustic soda to separate the iron 

 oxides and entangled matters ; the metallic portion con- 

 tained traces of cobalt, but the indications for nickel were 

 not satisfactory. The cobalt was present in less quantity 

 in this than in the other roof dusts. 



jRoof Dust. — In March 1882, Mr. Russell, Government 

 Astronomer, collected for me some dust from the roof of 

 the Sydney Observatory. From 493 grams of this the 

 magnet extracted 48*5 grams, or roughly 9*8°/>; but as this 

 evidently contained much entangled non-magnetic matter 

 it was boiled with caustic soda and repeatedly extracted 

 with a magnet under the soda solution. The proportion of 

 magnetic matter to the original dust was then found to be 

 only l'5 /°. Under the microscope it had the appearance 

 of Fe 3 4 , but with it were some lustrous steely looking 



