Tol | 
XVII. 
ON THE QUANTITATIVE SPECTRA OF IRON, ALUMINIUM, 
CHROMIUM, SILICON, ZINC, MANGANESE, NICKEL, 
AND COBALT. ; 
By JAMES H. POLLOK, D.Sc., anp A. G. G. LEONARD, A.R.C.S.I. 
[Puates XII.-XVII.] 
[Read Decemprn 18; Received for Publication, Drcrmprr 21, 1906 ; 
Published, Juny 22, 1907. ] 
INTRODUCTION. 
Tne following quantitative spectra of iron, aluminium, chromium, 
silicon, zinc, manganese, nickel, and cobalt, are, with slight 
modifications, taken after the manner devised by Professor 
W.N. Hartley, as published in the Philosophical Transactions of 
the Royal Society in 1884, vol. clxxv., pp. 49-62, and pp. 325-342. 
For analytical purposes a knowledge of the residuary lines of 
spectra is of the greatest assistance, as on diluting a solution of a 
salt, the lines of the spark-spectrum disappear so rapidly, that with 
0-1 per cent. solutions, the spectrum is difficult to identify by refe- 
rence to an index where all the lines of the element are carefully 
recorded, especially as the last lines that disappear are not necessarily 
the most intense. On a plate giving the spark-spectrum of some 
chloride of beryllium, let us say, one might very well have a few 
foreign lines closely agreeing with lines of iron, manganese, or 
titanium, and without very exact measurement, it might be im- 
possible to decide to which they belonged ; but on the other hand, if 
we knew that these lines were the residuary lines of titanium, and 
that the residuary lines of iron and manganese were quite different, 
we should know definitely that a trace of titanium was present, and 
not iron or manganese. Again, if we know that the residuary 
lines of an element are absent, we know at once that the element 
is absent, so that it is only necessary to look for the residuary lines. 
Hartley investigated the dilution spectra of a large number of ele- 
ments; but, with the exception of aluminium, silicon, and zinc, the 
SCIENT. PROC. R.D.S., VOL. XI., NO. XVII. 24K 
