114 CONTRIBUTIONS TO MINERALOGY. [bull. 262. 
B 2 3 . In 1887 Riggs gave an analysis of the New York mineral, in 
which he found 4 per cent B 2 3 . Two years later Whitfield gave 
several analyses of American dumortierite, all showing the presence 
of B 2 3 . In 1899 Linck gave some observations on the mineral, 
including an analysis by W. Schimpif , showing a strong test for boric 
acid. In 1902 Ford gave us three anatyses of American dumortierite 
and also mentioned two new localities. During the winter of 1903-4 
the writer made an analysis of the California dumortierite collected by 
himself during the previous summer. An analysis was also made at 
the same time of the Washington dumortierite kindly furnished by 
Mr. Brereton. 
Believing that one good analysis is better than several poorer ones 
for the exact determination of the formula of the mineral, there will! 
be presented an analysis of the California dumortierite made by the 
writer. From this a formula has been calculated for the mineral, 
and it will be shown how the other analyses agree with this proposed! 
formula. Before giving the results a few preliminary words will noti 
be out of place. 
An analysis of dumortierite is a difficult operation. The small 1 
amount of silica makes it difficult to get a good fusion. On prelimi- 
nary trials on an analysis of the California dumortierite, it was found 
that what was weighed as silica in the regular course of the analysis 
contained more or less of the undecomposed mineral. Results of 1 
about 30 to 32 per cent were obtained. This is, of course, on the 
uncorrected silica. It was found that a second fusion of the supposed 
silica was not only very beneficial, but also necessaiy. What was 
then weighed as silica contained but little residue. The handling of 
so large an amount of alumina is very cumbersome, and an accurate 
determination of the alumina (plus iron and other oxides here pre- 
cipitated) is difficult. The boric acid determination is tedious and 
difficult. The accurate determination of the water, which is given off 
only at a high temperature, is also not easy. One can thus see that 
an analysis of such a mineral is a troublesome operation and that a 
greater allowance must be made than for most silicate analyses. 
The specimen analyzed was selected in the field, an exceedingly 
pure piece weighing about 10 grams being chosen. The spec™ 
gravity of this piece was taken by suspension in water, giving 3.306. 
This was broken into small pieces and carefully examined for musco- 
vite, quartz, or other minerals. Sections of the mineral sh wed that 
the dumortierite was free from any inclusions. No grains of any tita- 
nium mineral abundant in the quartz could be detected. When the 
mineral was powdered it was treated with heavy solution, of specific 
gravity 3.10, and a minute amount of mineral (muscovite?) stayed on 
top and was removed. The separation was repeated several times, 
and the powder was dried at 100° and carefully examined under the 
