Abstract
During the course of a systematic programme of intensity measurements of band spectra it was considered advisable to include some measurements of the first negative nitrogen bands. These bands are emitted by the ionised nitrogen molecule and show very clearly the interesting phenomenon of alternating intensities. It was soon found, however, that the particular source employed possessed a number of useful properties, not least of which was its ability to excite in a selective manner the negative bands so that they were practically free from the usually troublesome second positive group. In addition, the bands were excited to such unusually high rotational levels that in this way a number of large and hitherto unobserved perturbations were revealed. It is the purpose of this paper to discuss these perturbations in detail; the other characteristics of this type of excitation will be discussed elsewhere.
Experimental.
The bands were excited in an ordinary Pointolite lamp, that is, in a comparatively low voltage (70 v.) tungsten arc in pure nitrogen at about 10 cm. pressure. Under these circumstances the 0, 0; λ 3914 band was obtained entirely unobscured, whilst the 0, 1; λ 4278 band contained but slight traces of the 1, 2; λ 4237 band. They were photographed in the second order of a 21-feet concave grating of the Physikalisches Institut, Bonn; in the case of λ 3914 two satisfactory plates being obtained with exposure times of 6 hours whilst for λ 4278 only one plate was obtained with an exposure of 10 hours. The plates were measured in the usual way, precautions being taken to eliminate any possible errors of the screw of the measuring micrometer. Intercomparison of the several sets of measurements (four) for each plate shows that the relative accuracy attained is about 0·06 cm.
-1
—this is the mean error for a large number of lines—whilst the two plates of λ 3914 indicate that for this band the absolute accuracy is also of the same order. For the purposes of this paper, however, the absolute accuracy is of lesser importance, since the information to be discussed is derived from measurements within a single band. Intensities were obtained from plates taken in the first order using the “raster" method of Frerichs. Since this paper is concerned primarily with the perturbations it will suffice to say that these plates were photometered with a Moll type recording microphotometer and the intensities evaluated from the density records in the usual manner. These intensity measurements will form the subject of a separate communication. A portion of the 0, 0 band is reproduced in Plate 21 with below it for comparison a photograph of the band as it appears when excited at the hollow cathode of a Geissler tube. The remarkable change in the appearance of the band is at once evident; the branches can be followed to about K = 80 (in a plate taken under smaller dispersion to K = 100) and this extension combined with freedom from extraneous lines makes the band under high dispersion a very fine example of its kind.