PHYSICS. 271 



slit and the prism-axis horizontally and projects tlie spectrum on the free 

 surface of the liquid. The author's spectroscope consists of a graduated 

 circle carrying the collimator and telescope, and movable to any angle 

 in the vertical plane. A mirror sends the light into the collimator, 

 beyond the lens of which and in the same tube is a direct- vision prism. 

 The spectrum is thrown on the surface of the liquid contained in a small 

 vessel on a movable table attached to the stand. It passes from it into 

 the observing telescope, whose focal distance is shortened by a second 

 object-glass removable at will. The circle is divided for the purpose of 

 measuriog the angles of incidence and emergence. — {J. FJujs.^ viii, 411, 

 December, 1879; Nature, xxi, 2G7, January, 1880.) 



Liveingand Dewarhave studied the conditions of reversal of the lines 

 of metallic vapors. They used at first a tube, the closed extremity of 

 which was heated to incandescence in a furnace. The metal contained in 

 the tube volatilizes; the light given by the bottom of the tube and by 

 its walls gives alone a continuous spectrum; but before reaching the 

 spectroscope it traverses the vapor. In the early experiments an iron 

 tube lo'^'" in diameter and 80'''" long was used, ])rotected externally by a 

 glaze. It was heated tor .'30^'", the upper extremity of the tube being 

 open and covered with a plate of mica or glass. A slow current of hydro- 

 gen ])revented the vapors in the upper end of the tube from oxidizing. 

 The collimator of the spectroscope had its axis in line with that of the 

 tube. In this way the green thallium line, the blue indium line, the 

 lines &, and some blue lines of magnesium were reversed. The temi)era- 

 ture not being sufticiently high, the authors next used a block of lin)e, 

 having a hole bored in its axis, (J or 7""" diameter, into the lower end 

 of which a lateral hole oi)ened for the insertion of an oxyhydrogen jet. 

 Or two lateral holes may be bored opposite to one another and the two 

 carbons of a voltaic arc introduced. Similar crucibles of carbon were 

 also used. By means of a mirror jilaced over the oi>ening the light was 

 reflected to the spectroscope. In this way the reversal was accom- 

 plished of 13 of the 31 lines of potassium, of 4 of the 12 of sodinm, of fi 

 out of 21 of barium, of 10 out of 34 of strontium, and of 11 out of 37 of 

 calcium. Beside these, 4 lines of lithium, 1 of ciesium, 5 of magnes- 

 ium, and 2 of aluminum were also reversed. — {J. J^hys., viii, 38.5, 

 November, 1870. From Proc. Boy. Soc, xxvii, 132, 3o0, 494, 1878: 

 xxviii, 352, 307, 471, 1880.) 



Peirce has called attention to the fact that a line in the diffraction spec- 

 trum (whether light or dark) must be shifted from its normal position 

 in case another line falls very near it. Neighboring lines must be at- 

 tracted if both are bright or both dark, and repelled if one is bright and 

 the other dark. The reason assigned is that the lines are only maxima 

 or minima of light, and the differential coefiScient of the sum does not 

 vanish at the same points as the differential coefficients of the separate 

 terms. The shifting will be the greatest in the case of a faint line near 

 a very intense one. The author has succeeded in shifting the position 



