MORSE. — SPECTRA OF WEHNELT. 527 



X 4352. The spectrum of the arc between magnesium terminals in an 

 atmosphere of hydrogen is nearly identical with that produced under 

 water.* 



3. The absence of certain strong lines in the Wehnelt spectrum. 

 Among these the arc line A 5529, the flame line A 4570, the spark line 

 A 3437.5. The presence of other lines usually appearing, and supposed 

 to correspond to the same temperature, is also of importance. 



4. The presence of the bands usually ascribed to the oxide. In the 

 baud with head at A 5006 seven flutings are visible. These are also 

 visible in the arc spectrum, but are much fainter. This band was de- 

 scribed by Liveing and Dewar f and resolved into fine lines by Crew 

 and Basquin. t The first-named authors obtained the band only in 

 oxygen, air, or carbon dioxide, and the evidence seems to show con- 

 clusively that it is due to the oxide. It seems clear from Liveing and 

 Dewar's work that the presence of this band does not necessarily indi- 

 cate a lower temperature than that corresponding to the other lines of 

 the spectrum of magnesium. This band also appears when the spark is 

 allowed to pass from a platinum point to a solution of a magnesium salt. § 



5. The appearance of certain lines. The line at A 4481 is perfectly 

 sharp, as it is when inductance is introduced into the circuit of the spark 

 in air. || The triplet A 3829.5, A 3832.5, A 3838.5 is very intense, and 

 the lines are broadened and diffuse in contrast to their appearance in air. 

 Wilsing If finds that when the arc passes between magnesium terminals 

 under water, this triplet is displaced, and broadens out into diffuse ab- 

 sorption bands. 



Calcium. 



The spectrum is very like that of calcium in the spark, with many 

 differences of intensity, and absence of some fairly strong lines. The 

 spectrum from A 5500 to A 5900 is of interest. There is evident diffuse- 

 ness in the lines in this region, in marked contrast to the sharp lines of 

 the spark and arc, but whether this is due to a real difference in the 

 spectra or a possible lack of adjustment of the apparatus a more ex- 

 truded investigation must decide. So far as the plates obtained show, 

 there are bands with heads corresponding to sharp lines in the arc and 

 spark spectra. 



* Astropliys. J., 17, 229 (1003). t L. c. 



t Astropliys. J., 2, 101 (1805). 

 § Vid. Lecoq de Boisbaudran, Spectres Lumineux. 

 || Vid. Huggins, Astropliys. J., 17 145 (190o). 

 IT Astropliys. J., 10, 118(1899). 



