ME. A. FOWLER ON THE SPECTRUM OF MAGNESIUM HYDRIDE. 457 



Moreover, the wave-lengths may readily be reconverted to the Fabry and Buisson 

 scale if so desired. 



The Intensity Scale. 



The intensities of the various lines have been roughly estimated on a scale of 10 

 for the strongest lines, ranging down to 1 for the fainter of the well-marked lines. 

 For the faintest lines 0, 00, and 000 have been employed, in accordance with the 

 convenient system adopted by ROWLAND ; these are only visible on the plates which 

 have received comparatively long exposures, and 000 is at the limit of visibility. 

 Many of the lines are probably composite, so that a line of stated intensity may 

 sometimes be made up of two or more lines of lower intensity. 







Series Lines in the Green Fluting. 



A preliminary analysis of the wave-lengths has been made for the purpose of 

 establishing series relationships between the lines of magnesium hydride, and thus 

 obtaining a general check on the accuracy of the wave-length determinations. Some 

 of the related lines were readily picked out by inspection, especially those proceeding 

 from the first heads of the green and yellow-green flutings. Others were identified 

 by their occurrence in pairs, and by the approximate constancy of the second 

 differences between the wave-lengths of successive lines of the series. 



The more striking of the probable series occurring in the green fluting are indicated 

 in the general wave-length table by the use of the letters a, ft, &c. , to mark the 

 associated lines. They are also shown graphically in Plate 12B. The yellow-green 

 fluting exhibits similar series, but the " stride " is considerably larger than in the 

 series composing the fluting in the green. The blue group of flutings is extremely 

 complicated, and no attempt has yet been made to resolve it into its component 

 series. 



Many lines which were measured and tabulated as single lines appear to belong to 

 two or more of the series, and it is, therefore, probable that such lines are very close 

 doublets or triplets which could not be separated with the resolving power of the 

 instrument employed. The same conclusion is also indicated by the greater intensities 

 of these lines as compared with adjacent members of the component series. On 

 account of this composite character of many of the lines, and partly for the lack of an 

 exact formula to connect the lines belonging to a given series, the series inquiry does 

 not furnish a complete test of the accuracy of the measurements. Nevertheless, 

 there is sufficient evidence to indicate that the values tabulated cannot be greatly in 

 error. 



The general structure of a series will be gathered from the following table, showing 

 the first and second differences between the wave-lengths of the lines composing the 

 a series : 



VOL. CCIX. A. 3 N 



