208 :" Trofs. Liveing and Bewar. Studies [Nov. 4, 



(2) Professors LiVEiNG and Dewar. Studies in spectrum 

 ancdysis. 



The authors describe the reversal of characteristic lines of 

 rubidium and coesium when the chlorides are heated with sodium 

 iu glass tubes in an atmosphere of hydrogen or nitrogen, and a 

 bright light is viewed through the vapours. They remark that 

 the violet lines of rubidium, and the most refrangible of the 

 coesium lines are first seen, and broaden out the most when the 

 temperature rises, contrary to what might have been expected from 

 the analogy of other cases. The absoi^ption lines observed coin- 

 cided with the bright lines of the metals heated in a tlame, not 

 with the lines which they give in a dense electric spark ; but the 

 authors obtained spectra similar to the flame spectra by passing 

 sparks from an induction coil, without a Leyden jar, between 

 beads of fused chlorides of those metals, although simpler spectra 

 were produced by the more abrupt discharges produced by inter- 

 posing a Leyden jar. The authors further described absorption 

 spectra produced by magnesium vapour when mixed with 

 hydrogen, potassium, and sodium respectively. That produced by 

 magnesium and hydrogen consisted of a line a little less refrangible 

 than the h group, and a band rather more refrangible than the h 

 group, fading away towards the blue. The constant appearance of 

 these absorptions when the vapour of magnesium in hydrogen was. 

 observed in a hot iron tube, led to the endeavour to obtain the 

 corresponding luminous spectrum. This they succeeded in doing 

 by taking sparks from an induction coil, without a Leyden jar, 

 between magnesium wires in a tube full of hydrogen. It appears 

 that the compound to which this spectrum is due is formed only 

 within a certain range of temperature, and is dissociated at higher 

 temperatures — for the spectrum is scarcely seen at all when a large 

 Leyden jar is used, which may be supposed to have the effect of 

 shortening the time of discharge and increasing the temperature. 

 Further, this compound does not seem to be formed when the 

 pressure of the hydrogen is much reduced. In the case of sodium 

 and magnesium they observed an absorption line in the green not 

 observed in either vapour separately; and when potassium and 

 magnesium were used, a characteristic pair of lines in the red 

 always appeared, and sometimes another line in the blue. The 

 authors have not yet seen these as bright lines. In the course of 

 observations on the spectra of sundry rarified gases the authors 

 have been led to conclude that electric sparks take a selective 

 course in a mixture of gases, and that the differences in the spectra 

 observed in different parts of the same tube are probably due to 

 the existence of more than one gas in the tube. Tubes of nitrogen 

 which did not shew the lines of hydrogen at all when sparks from 



