194 
PHYSICS: A. ST. JOHN 
apparatus already set up for another investigation was used, being kept cool 
by leaving the laboratory windows open. This method was very uncertain 
on account of the erratic weather and was otherwise unsatisfactory and was 
discarded after a single good photograph had been obtained. The spec- 
trometer system was then enclosed and the chamber kept cool by cans of ice 
and salt. By this means the temperature could be kept reasonably constant 
but it was found virtually impossible to mount and maintain a specimen 
long enough to get a satisfactory photograph, probably on account of the pres- 
ence of the salt vapor. Upon the recommendation of Prof. A. G. Webster a 
grant was made from the Rumford Fund of the American Academy of Arts 
and Sciences, Boston, in aid of the investigation which made it possible to 
install a small ammonia refrigerating machine loaned by the Automatic 
Refrigerating Company and to build a specially adapted spectrometer mounted 
in a well-insulated refrigerator box. With this equipment the temperature 
could be maintained indefinitely and there was no further trouble from melt- 
ing of specimens. A marked tendency to sublimation, however, was trouble- 
some until each specimen was mounted in a gelatine capsule when equilibrium 
was quickly established between the crystal and its vapor. Protected in 
this manner specimens were preserved for days. 
Commercial artificial ice was first investigated as it shows marked pris- 
matic structure. Unfortunately the prisms are distorted through pressure 
in the formation of the ice so that it is difficult to identify cleavage planes. 
Some photographs show spectral bands but no sharp lines upon which to 
base calculations. The investigation is to be pursued further, however, 
as it is probable that sufficiently small crystals or slices of crystals will give 
sharply defined lines. In a second procedure a thin layer of ice, about 2 
mm. thick, was allowed to form on a pan of tap water. It was difficult to 
identify individual crystals and more difficult to isolate them for mounting 
but occasionally a reasonably good specimen was secured and mounted so 
that the axis of rotation bore a definite relation to the original surface of the 
ice. Orientation with respect to other axes was a matter of guess work, 
usually wrong as the results showed. Nevertheless several satisfactory 
photographs were obtained from specimens prepared in this manner, in fact 
the calculations are based entirely on them. In a third method ice was frozen 
out of a weak salt solution. In this way a large crop of thin specimens show- 
ing distinct cleavage planes could be secured. Owing to the failure of the 
source of power the investigation was interrupted at this point and has not 
yet been renewed. One photograph was obtained showing distinct spec- 
tral bands but no identifiable lines, possibly on account of the mixture of 
microscopic salt crystals with the ice. The method is promising and is to 
be pursued further at a convenient season. 
Ice is commonly assigned to the hexagonal system of crystals 2 and is con- 
veniently referred to a triangular space lattice each cell of which has sides 
