A. W. Wright— Gases contained in Meteorites. 175 
As the greater breadth of the cometary bands indicates a 
density of the cometic gases greater than that in the tubes ex- 
mained, an experiment was made, as follows, for the purpose 
of observing the effect of increasing the density of the gas. A 
glass tube, having an internal caliber of about one centimeter 
and some twenty centimeters in length, was closed at one end, 
and through the sides were inserted two platinum wires, at 
are the middle of the tube, the inner ends of the wires 
eing In its axis and separated by an interval of about one cen- 
meter. Small fragments of the Kold Bokkeveld meteorite 
were dropped into the tube and shaken down into the closed 
end. The upper end was now drawn out to a narrow neck, and 
the whole attached to the pump. After exhausting the air, 
the neck was sealed, the tube withdrawn, and supported in a 
vertical position so that the interval between the wires was be- 
fore the slit of the spectroscope, the end containing the meteor- 
te being below. By means of wires connecting the platinum 
points with an induction coil, sparks were passed across the in- 
terval, and when the lower end of the tube was gently heated, 
the characteristic spectrum of the gas evolved became visible. 
t first it was very similar to that which had been observed 
Previously, but as the heat was increased, and the pressure of 
© gas became greater, the bands were seen to widen out, 
until they at length fully equalled in breadth those of the 
haan and finally they showed a tendency to run together. 
€ order of their relative intensity there was no appreciable 
ae slight disagreement in the positions of the first two 
fads with the reported observations of cometary spectra is 
y explained when we consider that for the latter a 
Point of maximum brightness removed toward the middle of 
of ath, “t 
: fag of the 
thes a edge of the slit. A simple experiment with 
. account for the apparent want of agreement 
iscrepancies ; the bands, and also to explain some of the 
és In the results of different observers, as to the posi- 
