IN RELATION TO QUANTITATIVE CHEMICAL ANALYSIS. 
51 
which increases with their increased refrangibility. They appear to be the spectrum 
of nitrogen of the first order. Messrs. Parry and Tucker’s photographs are 
representations of the less refrangible rays only, such as are transmitted by glass 
lenses and prisms. For obtaining sparks from solutions, M. Lecocq de Boisbaudran 
uses a concentrated solution contained in a glass tube, through the bottom of which 
passes a platinum wire which is fused into the glass. The wire is connected with 
the negative pole of an induction coil, the positive pole terminating in a wire placed 
conveniently above this. A well-known modification of this is the tube of MM. 
Delachanel and Mermet (Schutzenberger’s ‘ Traite de Chimie Generale/ vol. i., 
p. 284), but as the rays must necessarily pass through glass, it is not adapted to 
the examination of ultra-violet spectra. Any method is disadvantageous which 
necessitates the use of platinum electrodes, since the number of lines belonging to this 
element is embarrassing when they have to be eliminated from the spectra of other 
bodies. M. Lecocq de Boisbaudran states that iridium points give no lines 
belonging to that metal, but this statement does not apply to photographed spectra 
of rays of high refrangibility obtained with a condensed spark, since about 180 lines 
have been counted in such a spectrum. Of all metals affording materials for 
electrodes gold appears to be the best ; its spectrum is a weak one, containing 
comparatively few lines, it is an excellent conductor of electricity, and it is not 
attacked by solutions of metallic chlorides. 
There is no very simple method of obtaining spectra free from air lines without 
resorting to either an inconveniently short or a too prolonged exposure, and if such 
existed it would be of doubtful advantage, since the number of the air lines in the 
ultra-violet region is not exceedingly large, and from their character they are easily 
recognisable. They are, in fact, a positive advantage, since by their well known 
appearance and positions they serve to fix the position of other lines. A simple and 
elegant apparatus devised by my colleague, Professor O’Reilly, consists of a metallic 
wick composed of a few twists of platinum wire, by which a Bunsen flame is fed with 
the solution to be examined. Wicks of gold wire have been used by me, the metal 
projecting about one-eighth of an inch above the surface of the solution. 
Of all elementary substances hitherto photographed, graphite yields the simplest 
spectrum (“ Note on Certain Photographs of the Ultra-Violet Spectra of Elementary 
Bodies,” Journal of the Chemical Society, vol. xli., p. 90). It shows lines and the 
edges of. bands due to the air-spectrum to the number of sixty-six, of rays assignable 
to carbon it contains but twelve. Graphite being an excellent conductor of electricity, 
electrodes are made in the simplest possible manner by cutting plates of good Ceylon 
or Siberian graphite, tapering from one-fourth to one-eighth of a inch in breadth and 
three-eighths of an inch in length. Deep grooves or scratches are made in the sides of 
the lower electrode, which is inserted into the end of a small glass JJ tube containing 
the solution. The upper pole may be of metal or of graphite, either being fused 
into a glass tube or held in a screw clip. Platinum terminals from the coil are 
H 2 
