transajctions of section a. 557 



5. The lines of bands, at least those of cyanogen and of aluminium oxide, are 

 not appreciably shifted. 



6. The shifts of similar lines of a given element are to each other as the wave- 

 lengths of the lines themselves. 



7. Different series of lines (as described by Kayser and Runge) of a given ele- 

 ment are shifted to different extents. When reduced to the same wave-length 

 these shifts are to each other approximately as one to two to four for the 

 principal, first, and second subordinate series respectively. 



8. Similar lines of an element, though not belonging to a recognised series, are 

 shifted equally (when reduced to the same wave-length), but to a different extent 

 than are those unlike them. 



9. Shifts of similar lines of different substances are to each other, in most 

 cases, inversely as the absolute temperatures of the melting points of the sub- 

 stances that produce them, 



10. The shifts of similar lines of different elements are to each other approxi- 

 mately as the products of the coefficients of linear expansion and cube roots of the 

 atomic volumes of tlie respective elements (in the solid state) to which they are 

 due. 



11. Elements belonging to the same half of a Mendelejeff group give lines which 

 shift proportionately to the cube roots of their respective atomic weights. 



12. The lines produced by those substances which, in the solid form, have the 

 greatest coefficients of linear expansion have the greatest shifts. The converse is 

 a'so true. 



13. The shift of similar lines is a periodic function of atomic weight, and 

 consequently may be compared with any other property of the elements which 

 itself is a periodic function of their atomic weights. 



A portion of this investigation was conducted jointly with Dr. J. F. Mohler, 

 and the whole of it under the direction of Professor Rowland and Dr. Ames, 

 Directors of the Physical Laboratory of the Johns Hopkins University. 



11. An Experiment with a Bundle of Glass Plates. 

 \ By Professor Silvanus P. Thompson, F.R.S. 



\2. A Tangent Galvanometer. 

 By Professor Silvanus P. Thompson, F.R.S. 



13. On the Constitution of the Electric Spark. 

 By Arthur Schuster, F.R.S. 



If the spark of a Leyden jar discharge is examined by means of a spectroscope- 

 it is found that the metallic lines are not confined to the immediate neighbourhood 

 of the poles, but are sometimes seen several millimetres away from the electrode*, 

 from which tbey must have been projected with considerable velocity. 



How to measure the velocity of projection has always seemed to me to be a 

 problem of interest. Apart from the information a knowledge of that velocity 

 might give us concerning the mechanism of the spark discharge, it is not impossible 

 that light might be thrown on some important points in spectrum analysis 

 which are at present under discussion. Thus, for instance, if the speed with which 

 a molecule is pushed forward into the centre of the spark depends on molecular 

 weight, we might hope to separate from each other those lines of a spectrum which 

 belong to different molecular combinations. 



At various intervals during a number of years I had made unsuccessful 

 attempts to deal with this problem, when I became acquainted with the elegant 



