1873.] Light. 275 



holding the spectroscope ; b b, brass clamp and screws for fixing it in its posi- 

 tion, and for holding the wire ring, c c, to stretch the wire supports, d d, of 

 a black curtain for keeping out stray light from the side-hole of the spectro- 

 scope ; e e, photographed glass millimetre, or other scale of equal parts, 

 sliding between two upright mahogany cheeks,//, and lighted from behind 

 by the gas-jet, g g; h h, a tin screen to shade the flame; e e, the dark-glass 

 scale with transparent graduations at mm, seen in the spectroscope above 

 and below the spectrum ; the upper one can be shut off by a black card screen, 

 1 1, slipped down in front of the glass plate; and all but three scale-divisions 

 of the lower one, at any point, can be stopped out by a card slider, k k, slipping 

 through slits in the two side-cheeks of the frame, to confine the vision to the 

 graduations closest to any line. The glass plate is backed behind by a piece of 

 oiled paper gummed by its edges to the photographic plate. Prof. Herschel stated 

 that he has also made an addition to Browning's pocket microscope. He put the 

 thinnest possible film of mica on the glass plate which fitted the eye, and he 

 held it in place with a small india-rubber ring. That film of mica, by in- 

 equalities of its thickness, or other refracting properties which it possesses, 

 eclipses certain rays of particular refrangibility, and, according to the thick- 

 ness, it will eclipse more or fewer bands in the whole range of the spectro- 

 scope ; so that, looking through it, you get the spectroscope divided into com- 

 partments by dark bands. As these bands are not easily seen, he thought 

 he would bore a hole in the side of the spectroscope, and get a scale 

 reflected in its prisms ; and if this scale was to be a pocket one, he must, 

 in this hole, which was bored through, put a small lens, and fix the scale close 

 in front of that lens, for use as a scale of reference. The use of a spectroscope 

 resolves itself into placing every line according to its scale position on any 

 arbitrary scale, but as far as possible on the scale of what is called the 

 natural standard scale of wave-lengths. If the position of every line which 

 is mapped can be given in wave-length, its description is then intelligible 

 to everybody. The scale fitted to this spectroscope, of uniform intervals, 

 would enable spectra to be recognised ; but being a new instrument its indica- 

 tions first require to be reduced to some well-known standard. The sim- 

 plest way of recording them would be by wave-lengths. The positions of 

 successive lines, as seen in the common spectroscope, are not in the simple 

 proportions of their wave-lengths ; the blue lines are more spread out than 

 corresponds to the differences of their wave-lengths, and the red lines are nearer 

 together, while the wave-lengths in the latter part of the spectrum are far 

 apart; and therefore the question was how to pass from a uniform scale used 

 in a spectroscope, in the manner shown by Mr. Procter, to the scale of wave- 

 lengths. He had quite recently found that a uniform scale, used with any 

 spectroscope, like that divided in Mr. Procter's instrument into millimetres, 

 is approximately proportional to a scale of inverse fourth-powers of the wave- 

 length. If, for example, we take the readings of tie sodium line, and of any 

 other known line of the spectrum, on a scale of equal parts, and replace 

 them and all the other readings of the scale in proportion by the inverse fourth- 

 powers of the wave-length, we would then find that we can pass from the 

 uniform scale to the wave-length by taking the inverse (or reciprocal) of the 

 readings so replaced, and taking the fourth-root of that reciprocal to obtain 

 the wave-length. It so nearly was the case in all prisms of ordinarv dispersion, 

 as to present an almost accurate means of passing without trouble from the 

 uniform scale of the spectroscope to the wave-lengths ; and although it is 

 not quite true of the whole range of the spectroscope, yet if used between the 

 short interval of two neighbouring lines to find the wave-length of an 

 intermediate line, it will give the wave-length of that line directly from the 

 milimetre scale. 



The following is a description of Procter's direct-vision micrometer scale for 

 pocket spectroscopes. A a', small direct-vision spectroscope, b b', a parallel 

 brass tube (first slide of a miniature toy telescope) braced to the tube and 

 draw-tube of the spectroscope by the double rings of bent copper plate, 

 a a, b b (seen also in Fig. 2). The ring, a, slides on the main tube of the 



