458 TRANSACTIONS OF SECTION B, 
has investigated this matter pretty exhaustively," and has stiggested four 
different methods for preparing reliable Hg,SO,.* The author wishes to 
suggest a fifth method which he has used, and which for simplicity and 
reliability seems to be superior to any of them. 
There is no difficulty in purchasing Hg.SO, which is quite free from 
foreign metals. But these commercially pure preparations, taken as they 
are, are unsuitable for use in standard cells because they have usually been 
precipitated from the nitrate solution in the cold, and are therefore not well 
crystallised ; free nitrate and basic sulphate are also present. But if com- 
mercially pure Hg,SO, be heated together with a little pure Hg and dilute 
H.SO, for a day or so at 120° to 150° in either a sealed tube (1 to 2 milli- 
metre walled tubing) or bottle with wire-bound stopper, and occasionally 
agitated, a white crystalline preparation will be obtained even from a 
specimen of Hg,SO, which is initially discoloured almost black. On opening 
the tube or bottle it will usually be noticed that the pressure has risen a 
little, and that there is a trace of NO, formed ; a little care should therefore 
be employed if there is more than a trace of nitrate in the original Hg,S0,. 
But the Hg,SO, obtained in this way is itself free from nitrate, and especially 
if the H,SO, be filtered off and renewed during the heating process; it is 
also, of course, quite free from basic salt. It is filtered off, ground up in a 
mortar with one or two successive quantities of diluted H,SO,, and then 
with several quantities of saturated cadmium sulphate solution; it is 
filtered off in a Biichner funnel after each washing. In this way it is 
possible with very little trouble to prepare quite large quantities of very 
reliable, well crystallised, white Hg.SO, in every way suitable for standard 
cells. 
Heating with HCl and Hg is also a very convenient method of preparing 
crystalline calomel; but in this case the heating takes about twice as long 
owing to the small solubility of calomel compared with Hg,SO,. 
5. A New Method of producing a Cadmium Arc. 
By T. Martin Lowry, D.Sc. 
In order to produce a cadmium spectrum of sufficient intensity for polari- 
metric work advantage is taken of the favourable properties of the silver- 
cadmium alloys. On account of their isomorphism the two metals form an 
excellent series of alloys which are characterised by good mechanical pro- 
perties and very high melting points. (An alloy with 60 per cent. Cd melts 
as high as 700° C.) In striking contrast to the behavicur of the pure metal 
the alloy gives a steady are which can be kept true to centre by rotating the 
electrodes in opposite directions. The spectrum shows the silver as well as 
the cadmium lines, but these are so far separated that even with a low 
resolving power the slit of a spectroscope can be opened to its full width 
without any overlapping of the brilliant ‘blocks’ of light which take the 
place of the usual ‘ lines.’ 
6. Mercury and Cadmium Lines as Standards in Refractometry. 
By T. Martin Lowry, D.Sc. 
The series of wave-lengths commonly employed in refractometry— 
Ha Na HB Hy 
6560 5893 4861 4341 
red yellow blue violet 
is unsuited for general use in optical measurements on account of the 
1. E. Smith, B.A. Report, Cambridge, 1904, p. 33. 
? F. E, Smith, 2,A, Report, South Africa, 1905, p. 98. 
