﻿158 0. T. Sherman — Spectrum of Comet 0, 1886. 



successive spectra to be observed as far as the D line in the 

 fourth spectrum. The screw carries a rider bearing a double 

 recording apparatus. The instrument being then pointed to 

 the star, the observer brings portion after portion of the spec- 

 trum before the eye-piece. When he is fully convinced that a 

 patch of light is before him, it is brought up to a thick cross 

 wire, an automatic record made and the screw again turned till 

 the observer is again convinced that an image of the comet 

 is before him. Throughout the whole operation no light other 

 than that of the comet was admitted. The information to be 

 gained from the color of the image is, for faint light, very 

 slight, and save such information, the observer's whole knowl- 

 edge of his work is bounded by the fact that he started from 

 a certain spectrum-place and will be stopped when he reaches 

 a certain spectrum-place. Datum lines afforded by salted alco- 

 hol flame or Pliicker tube, are finally added to the record. 

 The record being withdrawn the observer has before him a 

 double series of positions. By means of the datum lines these 

 are converted into wave-lengths. If the same line is found in 

 four out of six spectra it is set down as observed, if less than 

 four, it is classed as suspected. If observed on every day of 

 observation the reality of the line is accepted, if missed on 

 some occasions it is only suspected. The faintness of the 

 comet's light renders these precautions necessary. 



Wave-length in ten millionth millimeters. 

 •6300 61 59 57 55 53 51 49 47 45 43 41 39 37 35 3300 



Spectrum of Comet C, 1886, and spectrum of low temperature Carbo-hydrogen. 



To such a process the spectrum in question was subjected on 

 May 26, 28, and June 4. The operation was repeated through 

 the first, second, and third spectra, and again through the 

 third, second, and first. The resulting loci have the following 

 approximate wave-lengths: 618*4, 600*6, 567*6, 553*7, 517*1, 

 468*3, 433*2, with strongly suspected loci at 545*4, 535*0, 412*9, 

 and 378*6. These positions may be in error by one or two 

 units. It is of interest to compare them with the following, 

 given by Hasselberg* as the chief lines of the Hydrocarbon 

 spectrum: 618*7, 599*7, 563*4, 553*9, 5472, 516*4 469*7, 431*1. 

 Of these loci 553*7, 517*1, 468'3, may well be the bands 

 common in cometary spectra; 567*6 not uncommon, while 

 618*4 and 433*2 are probably those suspected by at least two 



* TJeber Die Spectra der Cometer Dr. B. Hasselberg, p. 21. 



