296 FirELD CoLuMBIAN MuseEuM—GEOLoGy, VOL. I. 
and this applies to the chondri as well as to every other part of the 
stone. In this view I agree fully with Weinschenk as quoted above. 
An analysis of the meteorite was made by Mr. H. W. Nichols, 
Assistant Curator of the Department, in the laboratory of the 
Department. 
For this analysis a fragment of the meteorite free from visible 
oxidation was pulverized and dried at 100° C. | 
A portion of 3.3863 grams was weighed out for the major part of 
the analysis, experiment having shown that better results could be 
obtained from portions of this size than from a larger portion of 
about '16 grams. The nickel-iron was separated by Eggertz’s method 
of solution in iodine, the stone being found to be of too compact a nature 
to admit of magnetic separation even if the iodine method is not to 
be preferred in any case. The siliceous portion remaining was 
separated into two parts in the usual manner by treatment with 
dilute hydrochloric acid and potash. The separated portions were 
not weighed as it was found that sufficient oxidation occurred while 
burning off the filter to vitiate the results, but were analyzed 
separately and their weights calculated from the analyses. The 
insoluble portion was fused with sodium carbonate and a small 
amount of nitre. Silica was determined after the common method. 
Nickel and cobalt were separated from the iron by three precipita- 
tions and long digestion with ammonia and a large excess of 
ammonium chloride. Cobalt was separated from nickel by potassium 
nitrite. Nickel was titrated with KCn and cobalt weighed as sul- 
phate. Magnesium was weighed as pyrophosphate and calcium as 
oxide after an oxalate precipitation in the usual manner. Chromium 
was weighed as lead chromate after oxidation by bromine in acetic 
acid solution. Phosphorus was separated by the acetate process and 
weighed as molybdate. Water above 100°C. was determined in a 
separate portion by Penfield’s method of direct weight. Sulphur was 
determined in a separate portion as barium sulphate. Iron and insolu- 
ble alumina were determined in a separate portion, the iron being tit- 
rated by permanganate. The alumina was weighed directly, that of the 
soluble portion as alumina, that of the insoluble portion as phosphate. 
The alkales were determined in a separate portion after separation 
by platinic chloride as usual. They were found to occur wholly in 
the insoluble portion. 710, was present in distinct although 
unweighable quantities. A precipitate of ammonium manganese 
phosphate also proved to be not quite large enough to weigh. A 
search for copper gave only a very faint brownish coloration with 
hydrogen sulphide which was not sufficient to verify. 
i i i i 
