MR. A. FOWLEK ON THE SPECTRUM OF MAGNESIUM HYDRIDE. 467 



It will be seen that the solar wave-lengths do not show such uniform second 

 differences as are given by the magnesium hydride lines. Various small corrections 

 might be made to bring them into better accordance with the series law, but at least 

 one of them must be of the order of - 025 tenth-metre, which seems to be too great an 

 error to be attributed to ROWLAND'S relative wave-lengths in so small a region of the 



o o 



spectrum. The series evidence, so far as it goes, is accordingly somewhat against the 

 identification of solar lines with magnesium hydride. 



As a result of the whole investigation, it is only possible at present to say that a 

 very small percentage, if any, of the faint lines of the normal solar spectrum are likely 

 to be accounted for by magnesium hydride. 



Summary. 



(1) No sufficient reason has been found for modifying LFVKING and DEWAK'S 

 conclusion that the spectrum under investigation is produced by the combination of 

 magnesium with hydrogen, and the spectrum has accordingly been referred to 

 throughout as that of magnesium hydride. 



(2) The magnesium hydride spectrum exhibits lines at very short intervals from 

 the extreme red to X2300, and shows definite groups of flutings having their first 

 head lines in the yellow-green at 5621 - 57, in the green at 52 11 'II, in the blue at 

 4844-92, in the violet at 4371'8, and in the ultra-violet near 2430. 



(3) From photographs of the magnesium arc in hydrogen, taken with a concave 

 grating of 10 feet radius, the positions of close upon 2000 lines composing the three 

 principal flutings have been determined. The wave-lengths were derived from the 

 interference standards of FABRY and Buissox, but have been converted to Rowland's 

 scale to facilitate comparison with solar spectra. 



(4) Twelve of the series of lines which compose the green fluting have been traced, 

 and it is shown that none of the formulaa which have been proposed are sufficiently 

 general in their application to represent all of these series within the limits of error of 

 measurement. For the longer series the closest approximation is given by HALM'S 

 equation. 



(5) The identification of magnesium hydride in the sun-spot spectrum has been 

 fully confirmed, and is clearly demonstrated by photographs which are reproduced. 



(6) It is shown that many of the bright interruptions of the dark background of 

 the spot spectrum are not bright lines, but merely clear interspaces between lines or 

 groups of lines in the spectrum of magnesium hydride. 



(7) The presence of the magnesium hydride flutings, together with flutings of 

 titanium oxide and calcium hydride discovered at Mount Wilson, accords with the 

 view that spots are regions of reduced temperature, and that their darkness is at least 

 partly due to absorption. 



(8) The investigation of the possible presence of lines of magnesium hydride in the. 



3 O 2 



