460 ME. A. FOWLEE ON THE SPECTEUM OF MAGNESIUM HYDEIDE. 



senting the series by equations of so simple a type. The equation is especially 

 unsuited to the y series, assuming that the last lines have been correctly assigned ; 

 the second differences in this case vary rapidly towards the end, and only the first 

 nine lines are reasonably represented by the equation. 



The long series marked e and , which comprise many of the stronger lines of the 

 green fluting, provide a better test of formulae which have been proposed, and it may 

 be of interest to refer to some of the results obtained for the series, which includes 

 about forty lines extending over a region of about 160 tenth-metres. 



When calculated for the first thirteen lines of the e series, LESTER'S equation gives 

 no error as great as O'O! tenth-metre, but if eight additional lines are included, and 

 the two extreme lines are used in the calculation of constants, the maximum errors 

 exceed O'l tenth-metre, and the formula is clearly inapplicable to the whole series. 

 Better results have been obtained by including another term in the equation, but an 

 appreciable systematic error remains even if only lines between the first and twenty- 

 second are employed in the determination of constants. If applied to the whole 

 series, this extended form of LESTER'S equation gives systematic errors which reach a 

 maximum of O'l 5 tenth-metre. 



Another equation which has been tested on the e series is that deduced by HALM* 

 from one adapted to line series, namely, 



XD "~ X,,j 



where the constants X , a, /3 and /a, are calculated from four ol the lines. This 

 represents the first twenty-one lines with an accuracy nearly equal to that of the 

 measurements, but considerable systematic errors remain when it is applied to the 

 whole series. The true equation connecting the lines of a fluting series, therefore, 

 remains to be discovered, and until some progress has been made in this direction, 

 and still greater dispersion can be effectively employed, it does not seem worth while 

 to pursue the matter further. 



The numerical results of the calculations for the e series, however, may be useful 

 to those interested in this inquiry, and they are accordingly brought together in the 

 appended table. Numbers below the horizontal lines are results of extrapolation. 



Comparison ivith the Sun-spot Spectrum. 



It is interesting to recall that the possible local occurrence of magnesium hydride in 

 the sun was clearly recognised by LIVEING and DEWAR in their paper of 1880.t 

 Referring to the head of the green fluting at 5210, they wrote : " A line of the same 

 wave-length has been seen by YOUNG in the chromosphere once. Its absence from 

 the Fraunhofer lines leads to the inference that the temperature of the sun is too high 



* ' Trans. Boy. Soc. Edin.,' vol. 41, p. 551 (1905). 

 t 'Eoy. Soc. Proc.,' vol. 30, p. 97 (1880). 



