No. 4.) 



RESEARCHES ON METEORITES— MERRILL. 



9 



By recalculation the composition as a whole is found to be — 



Per cent. 



SiOj 34.235 



AljO 7. 107 



Cr,0, 0. 423 



PA 0. 387 



FeO 10. 535 



MnO 0. 289 



CaO 1. 159 



MgO 20. 958 



NiO 0. 563 



CoO 0. 058 



NajO 0.913 



KjO. 

 SO,. 



0.171 

 0.285 



Silicate 

 portion. 



Fe 15. 526] 



Ni 1. 196lMetallic 



Co 0. llSfportion. 



P 0. OllJ 



Fe 3.8801 



S 2.220 



■Troilite. 



Total 100. 031 



The analysis given by Tichomirow and Petrow is as follows : 



Per cent. 



SiOj 36. 36 



FeO 13. 80 



MgO... 18.54 



CaO 3.00 



Fe 19. 80 



Ni 5.55 



S 2. 30 



P 0.05 



C 0.08 



CuSn 0. 52 



100. 00 

 The discordance in these results is altogether too large to be accoimted for satisfactorily. 

 That there must have been some error in the percentage of nickel, as suspected, is evident, as 

 Dr. Whitfield's analysis of the metallic portion shows but 7.158 per cent of this constituent, 

 which, when calculated in percentage of the entire stone, amounts to but 1.196 per cent instead 

 of 5.55. The discrepancy in the calcium oxide (1.159 per cent against 3 per cent) is greater 

 than should exist in portions from the same mass, but, singularly enough, the amount of troQite 

 as indicated by the 2.30 per cent of sulphur is about the same and the remaining differences are 

 perhaps not greater than might be expected with the exception of the alkalies, Whitfield report- 

 ing 1.084 per cent. 



It is to be noted fm-ther that Whitfield reports no traces of tin or copper and that the siH- 

 cate portion freed from all metal by boiling the finely pulverized mineral in mercuric chloride 

 still yields 0.744 nickel and cobalt oxide. It may be recalled that in the table of analyses given 

 in the Memoirs of the Academy ' there are to be found several instances of this character. 



To these, at the time, I made no reference in the text, feeling that in some instances at 

 least they might be due to imperfect separation of the metal from the silicate portion. In 

 analyses since made especial care has been taken to guard against any such possibility and there 

 seems no reasonable doubt but that the sUicates-olivines or pjToxenes, or both — in meteorites 

 carry small quantities of these constituents, as is the case in terrestrial rocks. Such being the 

 case, it follows that the statement made by Dr. Prior,' together with an explanation by Dr. 

 Wahl ' to the effect that " the ferromagnesium minerals of chondritic stones contain practically 

 no oxide of nickel," is founded upon faulty analyses and insufficient data. 



' Vol. 14, 1916, pp. 7-27. 



> Min. Uag. Nov., 1916, p. 39. 



• Uin. u. Petr. Mitthell, vol. 26, 1907. 



