Apeil 17, 1908] 



SCIENCE 



613 



ity different from those vibrating perpen- 

 dicular to it. Kerr's law for the relative 

 retardation of two such vibrations is given 

 by the formula 



d = B VH/dr 



where V is the difference of potential be- 

 tween the two parallel plates of a plane 

 condenser producing the field, d their dis- 

 tance apart, and I the length of path of 

 the ray in the field. B is thus the constant 

 of the dielectric used. 



For different reasons it is important to 

 determine this constant for carbon bisul- 

 phide. Previous investigators have found 

 it to fluctuate over a range as high as 

 twenty per cent. This paper deals with 

 the treatment of the liquid necessary to 

 reproduce results, and finally gives a series 

 of careful determinations of B for various 

 wave-lengths. Incidentally the results of 

 Blaekwell, showing that Kerr's law for the 

 variation of the retardation with the wave- 

 length is in error, are supported. 



The Spectrum of Calcium: James Baenes, 

 Bryn Mawr College. (Read by title.) 

 The paper considers the changes pro- 

 duced in the intensity and distribution of 

 light in the lines of the calcium spectrum 

 obtained from an arc between metallic elec- 

 trodes in air at atmospheric pressure and 

 at lower pressures with varying current- 

 strength. The results can be explained as 

 a density effect rather than a temperature 

 one. 



The new triplets found by Saunders in 

 a copper arc moistened with CaCla appear 

 very clearly and sharp when the are is pro- 

 duced in a vacuum with a current of 12 

 amperes. 



Many attempts were made to obtain true 

 double reversals of the E and E lines with 

 a steady arc, but without success. False 

 multiple and double reversals appear in 

 some of the other lines. 



Upon the Magnetic Separation of the Spec- 

 tral Lines of Barium, Yttrium, Zirco- 

 nium and Osmium: B. E. Mooee, Uni- 

 versity of Nebraska. 



Preston observed a similarity in the 

 magnetic separation for spectral lines 

 which formed similar series. Runge and 

 Paschen confirmed this observation for a 

 number of substances. 



This research uses the magnetic separa- 

 tion of spectrum lines to search for series 

 in cases where they have not been before 

 observed. Runge used this method with 

 barium. 



The substances were volatilized by means 

 of a spark in a strong magnetic field (24,- 

 400 c.g.s. units per sq. em.) and photo- 

 graphed by means of a 21-foot grating. 

 The components were separated by a cal- 

 cite prism. Observations were made upon 

 barimm, yttrium, zirconium and osmium. 

 A table was presented, showing an ex- 

 tended comparison between the author's 

 results and some recently published by 

 Runge. 



The most prominent feature of the table 

 is the fact that most of the lines of the 

 substances here studied do not belong to 

 the types which are represented in the 

 well-known series groups. The next promi- 

 nent feature is that the types are seldom 

 duplicated. Thirdly, a great many types 

 have common intervals, whose difference 

 consists in a variety of the factors by which 

 the interval must be multiplied to produce 

 the actual separation. Fourthly, whether 

 or not the intervals are aliquot parts of a 

 normal interval a is not so decisive. It 

 can easily be contended that the intervals 

 are irrational fractions. 



Fifthly, the Zeemann triplets offer no 

 great advantages for the study of series 

 unless the magnitudes of the separations 

 separate into well-defined groups, as they 

 do not in zirconium, osmium and yttrium. 

 Sixthly, series are eliminated from the 



