DR. A. SCHUSTER ON THE SPECTRA OP METALLOIDS. 
43 
observers as a spectrum of oxygen ; but Angstrom, Wullner, and Salet have recog¬ 
nised it as identical with the spectrum of carbonic oxide. By taking the necessary 
precautions I have succeeded in tilling a great number of tubes, which show no trace of 
the presence of any carbon compound. I have used a Sprengel pump, with air trap, 
such as is sold by Mr. J. J. Hicks. The connexion between the pump and the 
vacuum tube is made by means of a ball and socket-joint. The fall tube of the pump 
ends at its upper extremity in a hemispherical socket. A hollow sphere is ground 
together with this socket, so as to fit perfectly into it. The inner part of the sphere 
communicates with the pump by means of a small hole, and ends at the other side in a 
glass tube. I have always fused my vacuum tubes directly to this tube. For ordinary 
purposes the sphere is slightly covered with grease and fitted into the socket. A small 
quantity of mercury poured on the top will then make the joint perfectly tight. 
I have nearly throughout the investigation avoided even this trace of grease. 
Mercury alone will run through the joint as the air is exhausted, but I find 
that strong sulphuric acid, although running through the joint, will do so only 
very slowly ; so that if care be taken to fill up the sulphuric acid from time to 
time, the joint will keep perfectly tight. Such a joint allows great freedom 
of motion, which is of great convenience, as the spectroscopic examination is 
always best carried on while the tube is attached to the pump. 
The vacuum tubes used were, as a ride, of the form given in the accom¬ 
panying woodcut (fig. 1); that is, they had the shape of an ordinary Plucker 
tube, to one end of which a tube of hard glass (a) was fused. On the other 
side it ended in a tube (b), which was fused to the air pump. The hard- 
glass tube was filled with some substance which gave off oxygen on heating. 
Permanganate of potash, chlorate of potash, and oxide of mercury were used, 
so as to eliminate any lines which might be due to the substance from which 
the oxygen was prepared. It is necessary to subject all these substances to 
a preliminary heating, in order to free them beforehand, as much as possible, 
from moisture, which always is attached to them. I find permanganate of 
potash to be the most convenient substance, and I have only used the others 
in test experiments. I have used capillary tubes, of different bores, and I 
think that some of the contradictory statements made by different observers 
are simply due to the fact that the term “ capillary tube ” is rather vague, and com¬ 
prises tubes of diameters sufficiently different to affect the spectrum. The dimensions 
of two kinds of tubes chiefly used are as follows :—- 
A 
B 
Distance between the poles 
. . . 14 
8 
Length of capillary tube 
. . . 4 
3'5 
Diameter of capillary tube 
. . . 0-07 
0T4 
The measures are given in centimetres. Tubes with the dimensions given under the 
heading A were generally employed. I have sometimes suppressed the capillary tube 
G 2 
