1903.] KOENIG—-ARTIFICIAL PRODUCTION OF CRYSTALS. 227 
manner as described under copper, but the filings only occupied the 
lower half of the tube. Upon the upper flat surface crystals form 
of the thick tabular type, the first pyramid prevailing over the basal 
plane. The crystals are coherent laterally, crustlike, over a loose 
aggregate of bright, light gray crystalline matter with indistinct 
faces. At the time I thought these two materials were alike. But 
recently, on re-examination, it is seen that whilst the crystals have 
become much tarnished, the gray material has not changed at all. 
The crystal layer was detached as much as possible from the loose 
substance, for the analysis, but it was not possible to do this 
thoroughly. 
The analysis of the crystals gave (0.216 gram) : 
Gui 66.37): 63) == 1. O05 27 
(Ni + Co) = 2.43: 58.6 = 0.0415 
PG EES Beiaio His = 0.4120 
1.0930 
99-70 
Ratio: 
(CuNi Co) : As = 2.655 : 1.000 
The analysis of the gray loose material gave (0.2325 gram): 
Cu = 44.30 :63 =0.7032 
Ni = 12.54 } : 58.6 = 0.2822 0.9854 
Co= 4.00 
AS = 39-25 275 = ee 
100.09 
Ratio: 
(CuNi Co) : As = 1.88: 1.00 = 2:1 
This then is typical Keweenawite, described recently by me 
(Amer. Journ. Sci., Vol. xiv, December, 1902) as found at the 
Mohawk mine. The non-tarnishing quality is inherent also in 
the natural mineral, as mentioned 7. c. The crystals on the other 
hand are mohawkite; the excess of arsenic making the ratio 
2.655: 1 instead of 3: 1 is explained by the impossibility of 
separating the crystals from the adhering keweenawite. 
Experiment of December 24, 1901.—Instead of filings, two frag- 
ments of the alloy were exposed in the incubator for twenty-seven 
hours. Hexagonal plates, very thin, formed upon a crust of gray 
material strongly crystalline. The plates stood at right angles to 
the surface and could be brushed off with small camel’s-hair brush. 
