92 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Historical Survey. 



Houston ^ who notes the changes which take place in the reversal 

 system as seen by an echelon when a zinc arc "begins to hiss," speaks 

 of the " striking forms of reversal," the distances between the different 

 components in the line varying in the different parts of the arc. 

 With one of his arcs and a small amount of vapor, he obtained the 

 three blue lines of zinc "without reversals." Under certain condi- 

 tions the three blue lines were "all doublets with components of 

 ec^ual intensity." 



Janicki ^ in his inaugural dissertation (1905) states that "an exami- 

 nation by the echelon of the lines of the zinc spectrum developed in a 

 capillary tube of 0.3 mm. diameter with external electrodes at a temp- 

 erature of about 460° showed them to be single lines." 



Nutting,^ in a paper on line structure, mentions the fact that Pliicker 

 tube spectra of rarefied gases moderately excited show narrow lines 

 of the simplest structure, " but with a heavy current or capacity in 

 parallel, if the pressure be greater than 3 or 4 mm. the lines broaden, 

 and finally, with a spark in series with the tube, widen into a continu- 

 ous spectrum, with the peculiar fluted appearance characteristic of 

 spark lines." 



He states further that "sparks between metallic electrodes give 

 lines far too broad for use as monochromatic sources. They are 

 never less than half a tenth-meter broad. The effect appears to 

 depend chiefly upon the amount of capacity used, and is greatly 

 heightened by the use of another spark in series; that is, it is due to 

 the steepness of the wave-front of the current wavc.^ Inductance weakens 

 the wings produced by capacity, and sometimes channels them, but 

 never reduces a line to a simple structure. Occasional lines will 

 appear to simply broaden out under the violence of the discharge, but 

 ordinarily it is simply a case of the dark background — between 

 spectra. of different order — becoming luminous." 



"Using a small current (0.02 amp.) of low voltage (5000) and low 

 frequency (60) and a minimum of capacity, and electrodes of iron 

 and brass, the spark lines were found to be still broad and diifuse. 

 Lines due to impurities (sodiiun, for example) occasionally appear 



5 Philosophical Magazine, 7, May (1904). 



6 See Annalen der Physik, 19, 36-79, Jan. (1906). 



7 Astrophysical Journal, 23. Xo. 1, Jan. (1906). 



8 The italics are the writer's. 



