KENT. — SPECTRA OF THE ZINC ARC AND SPAPK. 93 



fairly sharp on but a faint liackground, but a number of tests in- 

 dicated that it is impracticable to obtain narrow lines by introducing 

 impurities into the spark." 



Further, when discussing arc spectra in general, he writes: "The 

 structure which a line exhibits depends primarily upon its intensity; 

 that is, upon the amount of a substance vaporized and the intensity 

 of its excitation in the arc"; and specifically, in the case of zinc: 



" All four zinc lines are rather diffuse, and are usually found double 

 or triple.* * * The blue lines, 4810, 4722, 4680, are broad and diffuse, 

 and show a trace of structure on reversal." 



In a general discussion attention is called to the fact that the 

 structure of any one line is very variable, so much so that "we may 

 hardly speak of any line as having a fixed definite structure, even with 

 a minute specification of conditions of production." 



Types of lines are classified according to structure and behavior, 

 and the general conclusion drawn that to explain certain types — lines 

 which, when single, under some conditions become double or triple, 

 symmetrically or unsymmetrically, with receding components of 

 various relative intensities — the old absorption theory of reversal 

 is not satisfactory.^ 



In another paper ^° covering the results of a search for intense and 

 yet "pure" light standards. Nutting, sketching the development of 

 the typical normal line in either the open air arc or at pressures less 

 than atmospheric, states: — "with increase of intensity the line 

 broadens, and finally separates into two; * * * with further increase 

 the two components continually broaden and separate"; and of 

 highest "rank as to purity are the composite lines produced in the 

 vacuum tubes measured between extreme components." 



In a paper ^^ on relative intensities of spectrum lines an attempt is 

 made to show that the changes produced in spectra by varying current, 

 capacity, inductance, temperature and pressure, may be accounted 

 for by a single variable, or at most, two. He WTites : — 



" Several years ago the writer ^^ gave the steepness of the wave- 

 front through a gas as condition for the preponderance of the secondary 

 over the primary spectrum. Crew ^^ almost at the same time con- 



9 Nutting advances a theory of broadening, doubling and reversal in the 

 Astrophysical Journal of April (1906). 



10 Bulletin Bureau of Standards, 2, No. 3, Dec. (1906). 



11 Nutting, Astrophysical Journal, 28, 66 (1908). 



12 Astrophysical Journal, 20, 135 (1904). 



13 Ibid., 20, 284 (1904). 



