8 Annals of the South African Museum. 



up to the present been imagined, there being no reason to suppose 

 that the circumstances resulting from the compression of the air in 

 front of the meteorite and the rarefaction behind it during the 

 passage through the air were different in this case to that of any 

 others. Anyhow, the dehcate and unusually regular drift-effect 

 shows that the compressed air flowed uniformly in all directions 

 towards the back part in spite of the flat shape of the stone. 



In conclusion, the fact may be mentioned that the stone must 

 have had its present shape on entering the atmosphere, or must at 

 least have acquired it immediately afterwards, since a subsequent 

 change of form would have caused a tilting of the stone and a 

 brushing off of the drift-efi'ect. 



Part II. — Microscopic Structure and Chemical ComiJosition. 

 Bevised and annotated after the author's death. By Prof. 

 E. Klein, of Berlin. 



As a result of the unusual abundance of opaque ores, the sections are 

 not so transparent as are those of the other black chondritic meteorites 

 used for comparison ; and as much of the ore occurs either enclosed 

 in the silicates or as part of the chondrules and protected by the 

 shell, it was not possible to make the preparation sufiiciently clear 

 either by treatment with acid or neutral solutions of salts. The 

 microscopic examination is therefore exceedingly difificult and liable 

 to error, and for that reason it must not be concluded, because no 

 traces have been found of the accessory constituents so common 

 in chondritic meteorites — maskelynite, plagioclase,''' monoclinic 

 pyroxene and glass — that they are entirely wanting here. The only 

 substances which are undoubtedly present are rhombic pyroxene 

 (enstatite, from the chemical analysis), some olivine, nickel iron, and 

 iron pyrites, all of which together, in very varying quantities, make 

 up the chondrules. All the larger opaque particles are of nickel iron, 

 and can be easily determined both by their lustre in reflected light, 

 and by treating the slide with a solution of sulphate of copper. They 

 predominate in the matrix, but are rare in the chondrules, except 

 when they consist, as happens more frequently here than else- 

 where, entirely or very largely of nickel iron. In that case they 



* This has since been found in small quantities, and shows approximately the 

 properties characteristic of bytownite. A section almost normal to a gave an angle 

 of about 5T with the P cleavage trace. — C. K. 



