84 The Spectroscope in Mint Assaying. [January, 
both sections of the spectrum should appear of equal 
brightness, and, more especially, should begin to recede 
from the central line of the spectrum at the same moment, 
and should disappear at the same moment. 
The spectra being inevitable natural effeéts of physical 
causes, a variation between two specimens of supposed 
equal fineness would, in theory, be necessarily indicated by 
the respective lines failing to correspond in their reciprocal 
action. 
A serious source of error in these comparisons was soon 
discovered, viz., that if one electrode was nearer the centre 
of the slit of the spectroscope than the other, its spectral 
lines would appear proportionately longer than those of its 
vis-a-vis, even though both electrodes were of the same pure 
metal. It then became necessary to devise a special appa- 
ratus for manipulating the electrodes when under exami- 
nation. This was constructed by Mr. Samuel James, the 
machinist in the Mint, and admirably fulfilled its object. 
A woodcut of it is appended hereto. Its peculiarity con- 
sisted in an automatic combination of accurately propor- 
tioned screws, acting in opposite directions, by which a 
single motion of the hand sufficed to cause the upper and 
lower electrodes to approach or recede from the central line 
of contact in an equal degree. The electrodes, which con- 
sisted of small strips of metal cut to a point, were held by a 
suitable arrangement on the outer circumference of two 
metallic rings insulated from each other, the upper one 
slotted to receive a series of twelve electrodes of varying 
known fineness, and revolving horizontally, so that each 
electrode might in turn be adjusted to face a single elec- 
trode of unknown fineness fixed on the lower ring. Its 
object was to admit of the electrodes being separated to any 
desired extent, while preserving the line of vision through 
the spectroscope, directed to the centre of the spark. 
In regulating the height of the instrument, the apparatus 
was always adjusted by passing the spark between two 
electrodes of pure gold, so that, on separating the points, 
the respective spectral lines corresponded exactly in their 
reciprocal action. 
A systematic series of experiments was now commenced, 
in which the behaviour of the more volatile metals was at 
first studied, viz., lead, zinc, bismuth, tin, antimony, cad- 
mium, mercury, aluminum, &c. All these give more decided 
spectra than the less volatile precious metals, and some 
interesting results were noticed. Approximate illustrations 
of some of these spectra are appended. 
