The American Journal of Science/Series 3, Volume 22/Spectroscopic Observations upon the Comet b, 1881

XXV.—Spectroscopic Observations upon the Comet b, 1881; by Professor.

the Comet was brightest the weather at Princeton was very tantalizing. From June 25 to July 3, the comet was seen and observed on every night except June 30, and on none of them, except July 2, more than an hour at a time, the work being invariably interrupted by clouds or fog.

For the spectroscopic observations I have used both the one-prism instrument, by the Clarks, which belongs with the Equatorial, and the solar spectroscope by Grubb—the latter with dispersive powers varying, according to occasion, from two to six dense glass prisms. The telescope was the 9½ inch Equatorial.

The following are the principal facts made out so far:

(1.) The spectrum of the nucleus was found to be for the most part simply continuous; but on several occasions, especially June 25, July 1, and July 12, it showed distinct bands, coinciding with those of the spectrum of the coma. When brightest the spectrum could easily be followed from the neighborhood of B to a point well above G; and in the lower portion it showed color strongly.

(2.) The spectrum of one of the jets which issue from the nucleus was isolated on June 29th and found to be continuous. I think this was usually the case with the jets, but it is seldom possible to separate the spectrum of a jet from that of the nucleus sufficiently to be perfectly sure.

(3.) The spectrum of the tail appears to be a continuous spectrum overlaid by a banded spectrum, the same as that of the coma. The bands in the spectrum of the tail were followed to a distance of about 20′ from the head, on June 29 and July 1. The continuous spectrum ceased to be visible before the bands were entirely lost sight of, using a slit wide enough to unite the b 's into one band.

(4.) The spectrum of the coma shows only three bright bands with a faint continuous spectrum connecting them. No other bands could be found, though the continuous spectrum could be followed from about half way between C and D, to above G. The Fraunhofer lines could not be seen either in the spectra of the nucleus or of the coma.

While the comet was brightest, the bands, especially the upper and lower ones, were very ill-defined, so much so as to interfere with satisfactory measurements of position. After July 1 the definition became better.

(5.) The coma spectrum was very carefully compared with the spectrum of the Bunsen burner flame, with the spectra of Greissler tubes containing CO, CO₂ and ether vapor, and also with the spark spectrum of magnesium and air. The wave length of the less refrangible edges of each of the three bands was carefully determined by micrometer measures, on June 29, and on July 1, 2, 3, 6 and 12.

All the comparisons concur in showing a close, and so far as the dispersive power employed could decide, an exact agreement between the spectrum of the comet and that of the Bunsen flame. On the other hand the discordance between the comet-spectrum and the spectra of the Geissler tubes was striking. The lower of the three comet bands was the only one which was even approximately coincident with any band of the tube spectrum.

(6.) The measurements on the evenings named give the following numbers for the wave-lengths of the bands, viz:

The lower band was much the most difficult to deal with. The maximum of brightness seems to be, not at the edge of the band, but a little way up, and this perhaps may explain the fact that I obtained 5564 in the case of Hartwig's comet (while Von Konkoly obtained 5610—a much better result). Dr. Watts (Nature, vol. xx, page 28) gives 5634·7, 5165·3 and 4739·8 as the wave-lengths for the corresponding bands in the spectrum of the Bunsen flame.

(7.) The middle band, on June 29, July 1, 2, and 3, showed three fine, bright lines upon it, one just at the lower edge of the band, and the other two at distances of about 30 Ångstrom units—coinciding apparently with three lines which are seen in the Bunsen flame spectrum, though I did not succeed in measuring them.

It is hardly necessary to say that the evidence as to the identity of the flame and comet spectra is almost overwhelming; the peculiar ill-defined appearance of the cometary bands at the time of the comet's greatest brightness is, however, something which I have not yet succeeded in imitating with the flame spectrum. The comet spectrum on July 25th certainly presented a general appearance quite different from that of the later observations, as regards the definition of the bands.

Perhaps I may be allowed to record here a fact which has nothing to do with the comet, but was observed while adjusting the spectroscopes upon the sun in preparation for evening work. I find that the one-prism spectroscope shows the bright lines in the upper portion of the chromosphere spectrum, above h, better than any other instrument I have yet tried. I have hitherto always found it rather difficult to exhibit the two H's as bright lines to a person unused to the spectroscope, but with this instrument they are perfectly obvious—even obtrusive. The only (and indispensable) precaution needed is to put the slit accurately in the focal plane of the telescope for these special rays.

Princeton, July 14.