ABSTRACTS

A NEW TYPE OF MONOCHROMATOR

Lyot1 recently described a highly original monochromator, constructed by him for astronomical purposes. The operation of the instrument is based on the interference of light that has passed through a system of crystalline plane-parallel plates and crossed nicols. In contrast to ordinary slit monochromators, it makes it possible, like a filter, to view the whole object at once and at the same time separates only the required narrow line, only slightly weakening the total intensity of the transmitted light.

The instrument works in the following way. The light from the source successively passes through a series of plane-parallel plates cut from a crystal parallel to the optical axis, placed between crossed polarizers. The schematic arrangement is shown in Fig. 1. The principal sections

Fig. 1.

of the plates make an angle of 45° with the principal section of the nicols. Each subsequent plate is twice as thick as the preceding one. An oscillation with amplitude \(A\), having passed through the first nicol, is resolved in the first plate into two oscillations with amplitude \(A\sqrt{2}\) and with a certain mutual path difference \(\delta\). After the second nicol we have two rays (reduced to one plane) with amplitude

\[ \frac{A}{2} \]

and the same path difference; after the second plate and the third nicol, four components with amplitude

\[ \frac{A}{4} \]

and path differences \(0,\ \delta,\ 2\delta,\ 3\delta\), and after the last \((n+1)\)-th nicol \(2^n\) oscillations with amplitude

\[ \frac{A}{2^n} \]

and path differences \(0,\ \delta,\ 2\delta,\ 3\delta,\ldots,(2^n-1)\delta\). As a result of the interference of such a system of components, in the spectrum of the transmitted light there will remain only a series of separate bands of width

\[ \frac{2\lambda}{2^n}\,\delta, \]

separated from one another by distances

\[ \frac{\lambda}{\delta}. \]

The amplitude at the center of a band is equal to \(A\). Such bands are no longer difficult to separate by a simple light filter.

The action of the instrument can be made clear by considering the scheme of Fig. 2. One plate, oriented in the appropriate manner between two polarizers in transmitted light, gives a series of alternating bright and dark bands2 (Fig. 2a). The distance between the nearest points of equal brightness is equal to

\[ \frac{\lambda}{\delta}, \]

where \(\delta\) is the path difference created by the given plate. The second system \((N_2 P_2 N_3 (-))\), considered as

separately in an analogous manner, also gives a system of bands (Fig. 2b). As is clear from the diagram, the second system will be opaque for the bands transmitted by the first plate, 3 and 5 (the curves by no means need be added). The third system will remove the maxima 2 and 6, etc. It is not difficult to see that in the final spectrum the width of the remaining lines is determined by the thickest plate, and the distance between them by the thinnest plate. This also determines the choice of the thickness of the first plate and the number of all the plates.

Fig. 2.

Lio’s monochromator contains 9 plates of quartz and spar. To reduce losses due to the reflection of light, all the gaps between the separate parts of the instrument are filled with liquid immersion. Motion along the spectrum is effected by a slight rotation of the plates. The instrument operates in the interval between 3800 and 20,000 Å. The entire system is enclosed in a thermostat, since temperature fluctuations cause a displacement of the lines of 0.4 Å per hundredth of a degree. In the arrangement of the plates there are certain special subtleties, for a description of which we refer interested readers to the original article.

Fig. 3.

Such a monochromator will make it possible, for example, using the bright lines of the solar corona, to obtain detailed photographs of it without an eclipse of the Sun. It will make possible the study, visually or with the aid of accelerated cinematography, of the motion of the solar atmosphere: changes in faculae, flocculi, and protuberances.

In an article by K. Wild^3 a method of monochromatization is described that is similar to the method described above. Wild places between polarizing prisms

a quartz plate, cut perpendicular to the axis, and uses the rotatory dispersion of quartz. Preliminary monochromatization is achieved with a narrow light filter. The thickness of the plate is chosen so that the planes of polarization, set by the first nicol, are rotated for the limiting wavelengths of the transmitted region, \(\lambda_1\) and \(\lambda_2\), through an angle somewhat greater than \(180^\circ\). The second nicol extinguishes the regions adjacent to \(\lambda_1\) and \(\lambda_2\), and transmits a narrow band (Fig. 3).*

Here, obviously, one could also, following Lyot’s example, assemble an apparatus of several stages and obtain better monochromatization of the light. However, in the second method, in order to isolate fine lines one would have to use plates of too great a thickness, while constructing such a monochromator for isolating broad regions does not appear advantageous, except perhaps for the ultraviolet region, for which there are no filters. It should also be noted that both methods are suitable only for narrow beams (a few degrees) and therefore are applicable chiefly for astronomical purposes or for monochromatizing the light of a point source.

E. Brumberg

Literature

1. B. Lyot, C. R. 197, 1593, 1933.
2. Wood, Res. in Phys. Opt. 11, 166, 1919; Phil. Mag. 27, 1018, 1914.
3. K. Wild, Phys. Z. 35, 503—504, 1934.

* This method of monochromatization was used earlier in his work by S. I. Vavilov.

1. Lyot. ↩

2. Bright and dark bands. ↩