of Gases from Meteorie Iron. 295 
yielded by meteorites; and this investigation was undertaken 
in the hope that the spectroscope would reveal them if present, 
though their small amount or peculiar character should render 
their discovery by the ordinary chemical means difficult or 
impossible. 
The mode of obtaining the gas was similar to that pursued 
by Graham and Mallet, with the exception, that the iron, in- 
stead of being a single solid mass, was in the form of' fine 
chips, prepared by boring into the meteorite with a steel drill 
upon a lathe, those portions which came from near the exterior 
being rejected, with the observance of suitable precautions to 
prevent the admixture of foreign matter. The fine particles of 
the iron were placed in a thick tube of very hard glass, which 
had been carefully cleaned, and were pressed into it with a 
lass rod. It was found possible to bring the glass tube, thus 
partially filled, to a red heat, without its yielding very muc 
the atmospheric pressure. 
For the exhaustion a very efficient Sprengel pump was used, 
provided with a U-gauge, the limbs of which have an internal 
diameter of about six millimeters, and which is capable of giv- 
ing accurate readings to a fraction of a millimeter. The mer- 
cury is supplied to the fall-tube by a recurved tube, the 
branches of which are thirty-two inches long, an arrangement 
which effectually prevents the entrance of air. The pump is 
capable of carrying the exhaustion to such a point that it be- 
comes difficult to estimate the difference of level of the mercury 
surfaces in the gauge, and the electric discharge through a 
vacuum-tube exhausted by it becomes very feeble. That there 
was no leakage was shown by the fact that, after the exhaustion 
ad been carried to the extreme, the gauge maintained its posi- 
tion unchanged for days. 
To the exhaust-tube of the pump was attached a horizontal 
glass tube having two branches at right angles to its length. 
To one of these a vacuum-tube of the form ordinarily employed 
for spectroscopic work was attached, so as to have a vertical 
position, The other was closed by a stop-cock by which air or 
any gas could be admitted. The outer end of the horizontal 
tube was also provided with a stop-cock, and beyond this the 
tube containing the iron was firmly attached with cement. 
It was important to know what lines would appear in the 
spectrum from an ordinary tube exhausted by the pump, and 
containing air, in order to judge correctly as to the presence or 
absence of any gaseous bodies coming from the iron. It is 
well known that the lines of hydrogen, and bands due to car- 
n compounds,* appear when the air in the tube is rarefied to 
a high degree. A number of preliminary trials showed that 
* Pliicker and Hittorf. On the Spectra of Ignited Gases and Vapors, Phil. 
Royal Soc., vol. 155, p. 1. 
