SECTION III, 1883. NS 5] TRANS. Roy. Soc. CANADA. 
Note on Spectroscope Scales. 
By PROFESSOR E. J. CHAPMAN, Toronto. 
(Read May 25th, 1883.) 
Two methods, it is well known, are in use for the determination of the relative posi- 
tions of spectroscopic lines. By one method, the distances of these lines from a fixed point, 
or from one another, is measured on a graduated circle, angular measurements being thus 
obtained. By the other method—the only one that can be employed in the case of direct- 
vision spectroscopes, and one also in very common use in other cases—the distances are 
read off at once on a graduated scale, attached to the instrument and brought by a properly 
arranged prism into the field of view. It is to this latter mode of measurement that my 
present observations essentially refer. 
Unfortunately, scarcely two spectroscopes carry the same scale: nor, in the graduation 
of the scale, do they start from a common fixed point. Some scales commence at the A line 
of the solar spectrum, and others at the supposed commencement of the red end of the 
spectrum. In the latter case, the graduation not being uniform, the line A falls at 12 in 
some scales, and at 15, 17, etc., in others; and the line D, at 25, 26, 50, 60, and so on, aécord- 
ing, of course, to the relative closeness of the graduation. This is often a cause of uncer- 
tainty in the exact location of spectral lines, and especially of newly discovered lines, 
described by different observers. 
In Germany, spectroscopic observers are beginning to take the D line as a starting- 
point, this line corresponding, as well known, with the yellow sodium-line. Where, of 
course, as in powerful spectroscopes, the D line is split up, the edge of one of its separated lines 
may be taken. The divisions are then numbered from this line (as zero) in both directions : 
the minus sign (—) being placed before the numbers of the divisions lying between D and 
the red or least refrangible end of the spectrum, and the plus sign (+) before those which 
lie between D and the violet or most refrangible extremity. 
As the D or sodium line is so generally present in flame-spectra, and so readily defined, 
this plan is decidedly a step, but only a step, in the right direction. For the purpose of 
comparison, another fixed point is required. Ifa second fixed point of this kind were uni- 
versally accepted, a very simple calculation would bring all spectroscope scales into corres- 
pondence, and thus do away with all ambiguity, no matter what the actual graduation 
might be. With this view, I would suggest that the bright-red and always sharply defined 
lithium-line be made the point in question, and that the number of divisions between it and 
the sodium (D) line be assumed conventionally to equal 10 ; and, further, that the numerical 
position (so to say) of a spectroscopic line be always stated by comparison with this assump- 
tion. The position of the line on any scale can then at once be ascertained. For example, 
in a small (direct-vision) spectroscope that I often use, there are (in round numbers) just 18 
divisions between the sodium and lithium lines, and 34 divisions (very nearly) between 
