SIDEEOLITE. 63 



when the mass contains other minerals than the iron itself. Each separate 

 mineral is analyzed, but the general average of the whole mass is un- 

 known. Indeed, in many cases it may be difficult if not impossible, 

 owing to its structure, to obtain the composition of the whole, except 

 approximately. 



Since the same mineral has nearly the same constitution, whatever may 

 be its associations, the native iron would be expected to yield the same 

 constituents whether in a large mass alone or in minute grains associated 

 with other minerals. Consequently, the analy.ses of meteoric irons as now 

 carried on afford but little clue to their structure, for a pallasite yields the 

 same result as a siderolite, so far as the usual published analyses show. 



The specific gravity of the siderolites varies considerably, but by for the 

 greater portion lie between 7.50 and 7-90. The highest number is 8.31, and 

 the lowest 5.75, but the average may be considered to be about 7-70. The 

 analyses of the siderolites have been arranged according to their percentage 

 of metallic iron, for in that way their relations could be best shown ; but 

 where there are several analyses of the same rock, all are placed together 

 without regard to the percentage of the iron, except in the case of the first 

 one in the given series. The predominating percentages lie between 87.00 

 and 97.00, the average being about 91.00. The highest percentage of iron 

 is 99.81 and the lowest 37.00; but only seven analyses show a lower per- 

 centajje than 80.00. 



The percentage of nickel, as a rule, in the meteoric siderolites varies 

 inversely with the iron ; and if it is taken in connection with its associated 

 cobalt and the iron — whether free, or united with sulphur or phosphorus 

 — it is found that with but few exceptions the three elements make over 

 99 per cent of the mass. The nickel varies in amount from none, in seven 

 analyses, and to .10 and .20 per cent, up to 12, 14, and 15 per cent. In 

 three extreme cases it was found to be 24.708, 36.00, and 59.69 per cent. 

 The percentage of nickel in the Greenland siderolites is low. One serious 

 difficulty in estimating the relative proportions of the iron, nickel, and cobalt 

 is the apparent unreliability of many of tlie analyses — a difference of about 

 four per cent existing in the nickel as determined from the same siderolite ; 

 and in an extreme case one chemist obtained 14.70 per cent of nickel, and 

 another none. As a rule, cobalt is in amount less than one per cent, but in 

 a few cases reaches between two and three per cent, and in one instance is 

 between three and four per cent. Probably it is always present with nickel. 



