ME. A. FOWLER ON THE SPECTRUM OF MAGNESIUM HYDRIDE. 4G5 



Possible Indications of Magnesium Hydride in the Normal Solar Spectrum. 



According to the observations of YOUNG* and DuNER,t the " band lines " of the 

 sun-spot spectrum, many of which have now been shown to be due to magnesium 

 hydride, would appear to be present as very faint lines in the normal solar spectrum. 

 The same conclusion was also suggested by my own observations,^ which showed that 

 bright interspaces in the spot spectrum sometimes occupied spaces between nebulous 

 lines of low intensity tabulated by ROWLAND, the lines themselves not being seen 

 with the dispersion available. 



From observations made with the very powerful instruments at Princeton, however, 

 W. M. MITCHELL was led to express doubts as to whether the greater part of the 

 band lines are ordinarily exceedingly faint lines in the photospheric spectrum which 

 are brought into prominence by the vapours in the spot, and was inclined to the 

 opinion that the band lines are not present in the photospheric spectrum at all. In 

 opposition to this, HALE and ADAMS|| found, from measurements of their large-scale 

 photographs of the spot spectrum, that the wave-lengths of the band lines agreed 

 very closely with the wave-lengths of very faint lines in ROWLAND'S tables, and they 

 regarded this as a proof that the substance producing the band lines contributes 

 feebly to the ordinary solar spectrum. 



Nevertheless, a critical examination of the table given by HALE and ADAMS, taking 

 account of our present knowledge of the magnesium hydride spectrum, suggests that 

 many of the supposed coincidences are purely accidental. The strongest lines of 

 magnesium hydride, if present at all in the Fraunhofer spectrum, are only represented 

 by lines of intensity 000 or 0000 in ROWLAND'S tables, and it is accordingly very 

 improbable that the fainter lines would reveal their presence in any photographs of 

 the solar spectrum now available. Yet some of these faint magnesium hydride lines 

 appear to be identical with band lines then supposed by HALE and ADAMS to 

 correspond with solar lines. 



The chief difficulty in arriving at a satisfactory conclusion on this point arises from 

 the fact that there are so many lines in the magnesium hydride spectrum and in the 

 solar spectrum that numerous chance coincidences might be expected even if there 

 were no real relation between the two. The application of the theory of probabilities 

 is somewhat uncertain, and an attempt has therefore been made to obtain further 

 light on the subject by comparing the relative numbers of coincidences of strong and 

 faint lines of magnesium hydride with the solar lines. To allow for possible errors in 

 the two sets of measurements, and for the fact that ROWLAND'S standards have not 



* ' Amer. Jour, of Sci.,' 3rd series, vol. 26, p. 333 (1883). 

 t 'Recherches sur la Rotation du Soleil' (Upsala, 1891). 

 | 'Monthly Notices,' K.A.S., vol. 65, p. 516 (1905). 

 ' Astrophys. Jour.,' vol. 22, p. 4 (1905). 

 i| 'Astrophys. Jour.,' vol. 23, p. 34 (1906). 

 VOL. CCIX. A. 3 O 



