276 REPORT—1885. 
Report of the Committee, consisting of Professors ODLING, Hunt- 
INGTON, and HartLey, appointed to investigate by means of 
Photography the Ultra-Violet Spark Spectra emitted by Metallic 
Elements, and their combinations under varying conditions. 
Drawn wp by Professor W. N. Hartury, F.R.S. (Secretary.) 
Tue last Report of this Committee was presented at the Southport meet- 
ing of the British Association ; since then an investigation in detail has 
been prosecuted of the changes observable in photographs of the spectra 
of the metals when a series of solutions of definite strengths is examined. 
It had previously been shown that solutions containing the same element 
in different proportions emit variations of the same spectrum, the lines 
differing in number, length, and intensity ; and the converse—namely, 
that under the same spark conditions similar solutions of the same 
strength always emit the same spectrum. Furthermore, I have shown 
the invariable character of the cadmium, tin, lead, and magnesium lines by 
observations made on about five thousand photographs, including not 
fewer than two hundred examples of other metals. The reason of this 
arises from the fact that unless the spark be almost at the highest tempera- 
ture attainable, its emissive power is insufficient to affect the photographic 
plate in the usual period of exposure ; it follows from this that when a 
condenser of constant capacity is in circuit, variable conditions such as 
may be introduced by the electrodes being near together or far apart, or 
by the use of a large or small coil, do not affect the result. Sparks are 
shortened and the character of the spectra is greatly altered by the use 
of a coil with a stout secondary wire, an instrument introduced and 
employed by M. Eugéne Demargay. The use of an instrument of this 
kind is not well adapted to the photographic method of working, because 
the nature of the sparks is such that the graphite electrodes are rapidly 
burnt away and the sparks are very short. 
For the examination of solutions chlorides are generally employed, 
but sulphates and nitrates are also used. The electrodes are nearly 
always of graphite (‘ Phil. Trans.’ p. 52, Part I. 1884) ; sometimes gold,. 
copper, or platinum electrodes are required for special purposes, wires of 
the metal being twisted into wicks. 
The solutions examined generally contained 1 per cent., '5th, yy oth, 
and +,/;;th of metal. It is seldom that more than three or four lines: 
are visible in solutions of the latter dilution, and the rapidly diminishing 
number of lines in solutions weaker than ;),th per cent. is very striking. 
In the following tables the spectra corresponding to various solutions are 
given, and attention is particularly directed to the copper, silver, and tin 
spectra as illustrating this point. In many spectra it is impossible to. 
predict the line or lines which will be found to be the most persistent. 
Tt is also noticeable that the alteration of lines consequent on the dilution 
of solutions is variable in character with different lines in the same 
spectrum. Generally speaking, long lines shorten until they disappear, 
sometimes they become attenuated before they shorten, and in other cases. 
they attenuate until they fade away altogether. 
The calcium lines H and K attenuate considerably before they shorten,, 
while the lines of copper with wave-lengths 3273°2 and 3246°9, and of 
silver, 3382°3 and 3280-1, attenuate and fade almost away before: 
shortening. 
