complex spectra of stars of the solar type this is by no means the case; for, as Dr Hartmann remarks, “in the first place the lines in these spectra are so numerous that their complete measurement and reduction would require many days, and in the second place a rigorous reduction of such material has hitherto not been at all possible because the wave-lengths of the lines are not known with sufficient accuracy. On this account, observers have until now limited themselves to a partial treatment of such spectra, measuring only a small number of lines, whereby the major part of the rich material present in the plate remains unutilized.” But the spectroscopes that can be employed for stellar spectrographs are not sufficiently powerful to separate fully lines which are very closely adjacent, and therefore a line, assumed to be of a known wave-length, may be apparently displaced by the near neighbourhood of an unknown line. Hartmann overcame these and many other difficulties by directly superposing the image of the spectrogram of a star, having iron comparison lines, upon the image of a spectrogram of the sun taken also with iron comparison lines.
The apparatus for this purpose is shown in fig. 21, its principle of construction is shown in figs. 22 and 23. The solar spectrograph is attached by clamps to the plate A1, the stellar spectrograph to the plate A2. The plate A1 is mounted on the dove-tailed slide B1, upon the metallic stage T, and can be moved to right or left relative to T by the micrometer-screw S; whilst the plate A2 is mounted on the dove-tailed slide B1 and be moved at right angles to its greatest length by the screw G.
From Zeitschr, für Instrumentenkunde, by permission of Julius Springer, Berlin.
Fig. 21.
From Zeitschr, für Instrumentenkunde, by permission of Julius Springer, Berlin.
Fig. 22.
The micrometer-screw S has a pitch of 0·5 mm., its head is divided into 100 parts. Two spiral springs underneath press the plate B1 with its agate end-bearing against the rounded end of the screw S. The whole number of revolutions of the screw is read by the scale X (fig. 23). The whole stage T, carrying both spectrograms, can be moved from right to left on the steel cylinder Z, by turning the head K, on the axis of which is a pinion that gears into a toothed rack attached to the lower side of the cylinder Z. A scale N on the cylinder Z serves for setting the slide to any required position. The preliminary conditions of measurement are:—
1. The centre of both spectrographs shall be parallel to the axis of the cylinder Z.
2. The distance between the centres of the two spectrographs shall be equal to the distance between the optical axes of the two viewing microscopes.
3. The scales of the images formed in the focus of the eyepiece common to both microscopes shall be identical.
To fulfil condition (1) the plates A1 and A2 are mounted in circular slides, whose centres are E1 and E2 respectively, so that by means of the screws D1, D2, with their corresponding opposing springs F1 and F2, the operation can be very easily accomplished. To fulfil condition (2) the two microscopes whose object glasses are O1 and O2 (fig. 22) are attached to the plate L, their optical axes being normal to the stage T. The screw Q serves to adjust the axis of O1 to coincidence with the centre of the lines of the solar spectrograph, and the screw G then serves to move the slide B2 till the optical axis of O2 is coincident with the centre of the lines of the stellar spectrograph. Suppose now the solar spectrogram to be viewed in the focus of O1, and the converging rays to be reflected by the prisms P1 and P3, till an image is formed in the focus of the eyepiece at the point where the axis of the eyepiece intersects