368 RECORD OF SCIENCE FOR 1886. 



then reversed and ozone is passed into the apparatus to wash it out. 

 After more or less repetitions of this process, the hydrogen spectrum is 

 obtained pure. Tbe original paper must be consulted for the details. 

 (J. Phys., March, 188G, II, v, 100.) 



In a subsequent i)aper Cornu gives a description of the method em- 

 ployed by him for pliotographing the ultra-violet portion of the diffrac- 

 tion spectrum obtained from these hydrogen tubes, and also describes 

 the mode of measurement which he employed to obtain the wave lengths 

 of the lines from the photographic plate. Ten of the lines thus meas. 

 ured have wave lengths almost identical with those measured by Hug- 

 gins in the spectra of white stars. (J. Phys., August, 1886, II, v, 341.) 



In studying the photographic spectrum of hydrogen, Cornu has ob- 

 served that those groups of metallic lines which reappear periodically 

 with a particular regularity belong to the category of those which re- 

 verse themselves. Moreover, they get nearer together, and diminish in 

 intensity toward the more refrangible end of the spectrum. The spec- 

 tra of aluminum and thallium are excellent examples. Xow, in these 

 metallic spectra certain series of lines spontaneously reversed, present 

 sensibly the same law of distribution and intensity as that of the hydro- 

 gen lines. (J. Phys., March, 1886, II, v, 93 ; Nature, June, 1886, xxxiv, 

 105.) 



Wiedemann has cautioned experimenters engaged in studying spark 

 spectra against using for this purpose the spark of an induction coil, 

 becaut?e of its composite character. The spectra produced by such a 

 spark are multiple, their character changing with each of the partial dis- 

 charges; so that one of these discharges may produce a line spectrum 

 and another a baud spectrum. Moreover, it is not the sum of these im- 

 pressions that is observed, since the diminution of visual sensitiveness 

 with time is felt in a way which is quite different for these sparks of dif- 

 fering intensity. Hence he prefers to use the Holtz spark for spectra. 

 (Ann. Chim. Phys., January, 1886, VI, vii, 143.) 



Bell, under Rowland's direction, has determined the wave lengths of 

 thirty of the lines in the ultra-violet spectrum of cadmium. Seven of 

 the lines measured were in the visible spectrum, and their wave lengths 

 were determined by direct micrometric measurement. The others were 

 fixed from measurements of their photographs. The author believes 

 them correct to one fifty- thousandth part. (Am. J. Sci., 1886, III, xxxi, 

 426.) 



Mac6 de Lepinay has sought to determine the absolute wave length 

 of the line D2 by an ingenious and novel method, which consists in de- 

 termining optically, by means of Talbot's bauds, the dimensions of a 

 quartz cube as a function of this wave length, and thence its volume 

 in terms of this wave length, on the one hand. And on the other, in 

 determining by the balance, from the loss of weight in water, its abso- 

 lute volume in milliliters. The values obtained for the D2 line are, in 

 vacuo^ 5.8917x10* (milliliters) i; and la air 5.8900 xlO^C milliliters) i. As 



