On the Production of Spectra by the Oxyhydrogen Flame. 67 



configuration a function of h and v, the function ^ is still an 

 exact differential. 



Let S 2% or -£faf -J^ 8v be the new alteration in X , which 



is not the equivalent of work done, and let 8 2 x be the mean 

 value of 8^ 



Then, as the alteration is supposed to take place without ex- 

 pense to SQ 1? 8c,x must be subtracted from the whole change 

 of the ergal in the expression for 8Q. The equation (A) then 

 becomes 



Under the circumstances (B) becomes 



* v dv dh dv 



and by substitution we obtain, as before, 



Y B H — XlogA+ 3^%+ g w >=8(f). 



IX. On the Production of Spectra by the Oxyhydrogen Flame. 

 By Tasker H. Marvin, Brooklyn, N. Y* 



SIX or seven months since, in using the lime-light for spec- 

 troscopic work, I had the good fortune to discover a 

 method by which bright-line spectra, without an accompanying 

 continuous spectrum, could be produced at will by means of 

 the oxyhydrogen flame. 



I had been engaged in examining the spectrum of calcium 

 superimposed upon the continuous spectrum, which is easily 

 seen by bringing the collimator of a compound spectroscope 

 to bear laterally upon the incandescent point of the line. I 

 found that I could bring out all the bands of calcium and the 

 sodium line doubled sharply and brightly, notwithstanding the 

 brightness of the continuous spectrum. When, however, the 

 proportions of hydrogen and oxygen were so adjusted that 

 there was no excess of the latter, the bright lines of calcium 

 and sodium disappeared and the continuous spectrum alone 

 remained. The chemical reaction then seemed evident ; the 



* Communicated by the Author. 

 F2 



