PROCEEDINGS OF SECTION B. 395 
8.—ON THE PURIFICATION OF CERTAIN 
SUBSTANCES. 
By R. Turetratt, Professor of Physics, University of Sydney. 
9.—NOTES ON THE SPECTRA OF ZINC AND 
CADMIUM. 
By J. B. Kirxianp, F.C.8., Assistant Lecturer and Demon- 
strator of Chemistry, University of Melbourne. 
THE author, in the course of a spectroscopic examination of some 
zinc sulphide, obtained under peculiar conditions, was led to 
make a series of comparative quantitative spectroscopic experi- 
ments with samples of zinc obtained from different sources, in 
order to determine whether, under such circumstances, any 
difference in the sensitiveness of the spectral lines of zinc could 
be detected. 
Experiments were first made on a series of solutions of definite 
strengths, containing from 20 to ‘001 per cent of the metal as 
chloride, the apparatus employed consisting of a one prism 
spectroscope by Browning, and a spark apparatus of simple 
construction, the spark being produced by a 4-inch induction coil, 
excited by six large bichromate cells. 
The following results were obtained :—With the stronger 
solutions the spectrum consisted of four brilliant continuous 
lines, one in the red, w./., 6362°5, and three in the greenish-blue, 
w./., 4809°7, 4721°4., 4679:5.* As the solutions become more 
dilute, the lines gradually shorten to the negative electrode, 
until the solution contains ‘005 per cent. of zinc. With this 
degree of dilution the red line is not visible. Finally, with -001 
per cent., the others appear as dots on the edge of the spectrum. 
On repeating the above experiments, and introducing a con- 
denser (300 square inches of surface) into the secondary circuit, 
a pair of very bright lines in the green, w./,-4923°8, 4911-2, 
appeared, besides those already noted. These differ in character 
from the others, being blurred at the edges, and resemble air-lines 
seen in this region of the spectrum; moreover, they are not 
nearly so sensitive, for solutions containing one per cent. of the 
metal give them only faintly, and ‘1 per cent. almost 
imperceptibly. The other samples of zinc experimented on gave 
exactly similar results. ‘The metal itself was then submitted to 
the action of the induction spark. The spectrum was essentially 
the same as that produced by a strong solution of zinc, not a 
trace of the green lines being visible, even on increasing the’ 
* Wave lengths are given in ten-millionths of a mn. after Thalen. 
