110 FIELD COLUMBIAN MusEuM—- GEOLOGY, VOL. III. “ 
is that two ataxites fell at different times at one locality. In view 
of the small number of ataxites known, this seems less likely than 
to suppose that two masses of the same fall differed in composition. 
No other case of such marked difference is known. Differences of 
structure seem as a rule to be a better criterion for distinguishing mete- 
orites than differences of composition. On the other hand, similarity 
of structure and composition together do not positively .identify 
meteorites found at different places as belonging to one fall, since 
such similarities occur in meteorites seen to fall at widely different 
times and places. Of the nine iron meteorites seen to fall, four are 
medium octahedrites and have practically similar compositions. In 
correlating individual meteorites, therefore, all possible characters 
must be taken into consideration, including the circumstances of 
their find, the appearance of their exterior, the probable time elapsed 
since their fall;:etc. 
No attempt has been made by the writer at summation of the 
analyses here given, in order to determine the average composition of 
iron meteorites. Such a summation, if worthy of being performed at 
all, will be deferred until analyses of the iron-stone and stone meteor- 
ites are also at hand for comparison. ‘This work the writer hopes to 
accomplish in the near future. It 1s obvious, however, from an inspec- 
tion of the tables that the average percentage of iron in iron meteor- 
ites as a whole is not far from gi per cent, while that of nickel closely 
approximates 7.50 per cent. It is doubtful if the average percentage 
of the remaining minor constituents can be learned by summation of 
existing analyses. Not only have these constituents in many cases 
not been determined, but also any slight error in analyses or sampling 
would double or multiple their percentage. A percentage of .4 of 
cobalt, for instance, as compared with .2, is within the limits of error 
of many analyses, yet one percentage is double that of the other. The 
same 1s true in much greater degree of determinations of the amount 
of copper and other constituents. Until a larger number of complete 
and accurate determinations are at hand, therefore, summations of 
these constituents seem to have little value. One point in the compo- 
sition of iron meteorites which may or may not be of significance may 
be noted. Of the four constant metallic constituents, the most 
abundant, iron, has the lowest atomic weight, the next in quantity, 
nickel, is next higher, and so on for cobalt and copper. This grada- 
tion, using percentages common in iron meteorites, appears as follows: 
lron. Nickel. Cobalt. Copper. 
Per cent in iron meteorites- ---go 9 0.9 0.02 
Atomic=weight 22 oe 55-5 58.3 58.6 63.1 
