KENT. — SPECTRA OF THE ZINC ARC AND SPARK. 107 



by a Nernst lamp, an arc of great vapor density, and a highly disrup- 

 tive spark between terminals of the pure metal. ^^ These sources give, 

 in fact, nothing but the so-called " diflFraction " as distinguished from 

 the "interference" pattern. 



3. That inductance, even in small amounts, is extremely efficient 

 in reducing the intensity of the continuous or diffraction pattern and 

 producing structure in the spark image. 



4. That the structure varies with the part of the inductance spark 

 image used whether end or center; the end showing an enhancement 

 of the intensity of the components lying toward the red. 



5. That as the value of the inductance is increased, the red com- 

 ponents in the structure become less intense. 



6. That even a disruptive or non-inductance spark between brass 

 terminals shows structure in the zinc lines studied and that, if in 

 addition inductance be inserted, the resultant lines are as sharp, or 

 even sharper, than those given by a low current arc. 



7. That a small amount of vapor in the arc, even with fairly high 

 current (e. g. 8 amp.) produces conditions favorable to structure other 

 than the fluting which occurs when the arc is heavily charged with 

 vapor and is noisy. 



8. That on all plates obtained upon which the positions of the 

 components of the spark with small inductance are compared with the 

 positions of the components of the arc at low current (about 3.3 amp.) 

 the center of gravity of the spark structure lies further toward the red 

 than that of the arc. 



General Conclusions. 



That conflicting results were obtained by Janicki and Nutting is 

 probably due to the fact that different sources of light were employed. 

 The structure Nutting describes is unquestionably real. Certainly 

 echelon gratings may give ghosts. That the Petitdidier instrument 

 used in this investigation is free from such, is shown by the fact that 

 the green line of mercury shows no false lines. 



Further, from the visual observations made upon arc lines, it is 

 perfectly clear that the "ghost" argument will not explain the en- 

 durance of a satellite or its increase in intensity, when a formerly 

 brighter line grows fainter or disappears entirely, nor, specifically, 



21 This is true of the spark only when the echelon is not powerful enough 

 to resolve the components of the fluting. 



