October 12, ia^;5.] 



SCIENCE. 



489 



former, in 1871, in his work in mapping the 

 atmosphere lines, used for eoini)arison lilack 

 lines of various wiilths, ruled on white paper, 

 and viewed through vessels filled with dark- 

 ened water.' Gou}' made some measure- 

 ments of solar lines by photometric metliods ; 

 isolating a narrow strip of tiie spectrum adja- 

 cent to tlie line, and comparing its light with 

 that of a strip of equal width containing the 

 line. From these data he calculated the in- 

 tensity of the line, not in photometric, Init in 

 linear units.- The method adopted by the 

 writer is entirely ditfercnt from either of tliese ; 

 and. as far as known, is new. 



M'hat was desired was the production of an 

 artificial absorption spectrum, the intensity of 

 whose lines could be varied at will, until one 

 of the lines thus produced should be sen- 

 sibly tlie same as the line to be measured. 

 Fig. 1 is a section of the attachment to the 

 spectroscope made for this purpose. The 

 dark lines required are diffraction fringes i)ro- 

 dnced at the focus of the positive e^'e-jjiece, 

 which are therefore seen projected on the 

 spectrum. They are produced by placing a 

 silk fibre a little beyond the focus of the eye- 

 piece. In the figure, the piece a slides in tlie 

 tube, bearing with it a single silk fibre placed 

 vertically and just in the middle of the field of 

 view of tlie ej'e-piece. The fibre is maintained 

 vertical by means of a pro- 

 jecting pin sliding in a longi- 

 tudinal slot in the tube, as 

 shown at s in fig. 2. The 

 sliding motion is given to it 

 by means of the (liece 6. which '■ "■ 



turns freely, but cannot slide, 

 being retained by the screw d fitting in a 

 groove made entirely around the piece. Two 

 openings arc made in the tube, on opposite 

 sides, so that b can be turned directly with 

 the fingers. One of these windows is shown 

 at n in fig. 2. By turning in one direc- 

 tion, the silk fibre maj' be put nearly in the 

 focus : by turning back, it can be m.ide invisi- 

 ble. When near the focus, the fibre appears 

 as a pair of dark parallel lines and quite close 

 together. As it is drawn away from the focus, 



' .-Inn, pftj/a. cfihn.t xxiii. 274. 



' Complri rrndut, Ixxxix. 1033 and xci. 393. 



the lines appear to separate somewhat, grow- 

 ing constantly fainter until they disappear. 

 The fainter ditlVaclion fringes produced are 

 invisible in the rather weak light of the spec- 

 trum. Whole revolutions of the screw 6 are 

 read off on the graduation at the side of the 

 slot, and fractions (tenths) are read from the 

 piece itself, which is graduated as a microme- 

 ter screw. The lines thus produced resemble 

 closely the D group, particularly when both 

 are strong, when a very sharp eye is required 

 to distinguish the spurious lines from the genu- 

 ine. As the movement of the eye from side to 

 side would modify the appearance of the inter- 

 ference lines, making one darker than the 

 other, the spectrum must be viewed through a 

 narrow vertical opening, making such motion 

 impossible. For this purpose a piece of black 

 paper. (not shown in the figure), provided with 

 a vertical slit of perhaps 0.7 mm. width, must 

 be placed on the eye-lens at k. Even with 

 this, a little care is necessary in the position of 

 the eye, that the pair of lines shall always be 

 equal. The slight darkening of the spectrum 

 between the two lines, which occurs, is in this 

 case not objectionable, as it imitates prettj* 

 closel}' the general absorption in the space be- 

 tween D and the a line of the D group. The 

 instrument, as figured, is provided with a 

 tangent screw at e, bv which the whole tube 

 containing the eye-piece can be moved hori- 

 zontally, thus shifting the field of view so that 

 any line of the spectrum can be brought to the 

 side of the comparison lines. The instrument 

 is mounted on a wooden base, grooved at the 

 top to receive, it. At the back side is a large 

 knob bj^ which the instrument is held when 

 taking an observation. When directed to any 



part of the sky. tlie altitude can be determined 

 liy means of the graduated circle and hanging 

 weight shown in the figure. 



Another device, much simpler, and of use. 



