200 Absorption-Spectrum of Potassium Vapour. 



All the edges of the bands for sodium, that appear in the 

 photograph accompanying the paper of Wood's to which 

 reference lias already been made, are included in the above 

 table. It is seen at once that there is a definite corres- 

 pondence between the two cases. The actual closeness of 

 pairs of numbers in the two cases may not indicate anything, 

 but the fact that all the numbers for potassium fall in the 

 region occupied by the numbers for sodium is of interest, and 

 also the further fact that the spacing of the numbers is also 

 roughly the same. This seems clear evidence that the 

 mechanism at the back of the two cases is similar. When 

 one remembers the changes in this type of absorption spec- 

 trum produced by circumstances apparently trifling it seems 

 exceedingly probable that on fuller investigation the corres- 

 pondencies between the two spectra may be much more 

 definite than is at present observed. Much more detailed 

 observation is necessary in the case of potassium, and it is to 

 be desired that someone with the instruments necessary may 

 be able to make these observations. It seems clear, however, 

 that the two absorption spectra are analogous and that they 

 are connected with the first members of the principal 

 series. 



Anomalous Dispersion, by Potassium Vapour. 



The author has shown * that anomalous dispersion takes 

 place in the region of the first members of the principal 

 series in lithium and potassium, and further experiments 

 have been made with potassium to investigate other lines of 

 the same series. Anomalous dispersion has been detected at 

 the first four lines of the principal series for potassium, but it 

 is very small in amount after the second line 4047. The 

 effect can be easily detected on the negative, but as it is very 

 small excepting at the first two members of the principal series 

 it does not show sufficiently for a reproduction. The source 

 of light was a fine slit illuminated by an arc light. One 

 difficulty in the investigation is concerned with focussing 

 the image of the horizontal slit on the vertical slit of the 

 spectrograph. The focus cannot be good for more than one 

 line at a time as a quartz lens must be used, but further 

 experiments are to be made for this effect. 



I desire to express my thanks to the Grovernment Grant 

 Committee of the ]{o} r al Society for a grant which enabled 

 me to obtain the quartz spectrograph used for the work of 

 this paper. 



* Phil. Mag., Sept. 1909. 



