KENT. — SPECTRA OF THE ZINC ARC AND SPARK. 99 



General Method of Procedure. 



Before describing the work in detail it may be stated that the 

 general procedure was to set the echelon at the position of greatest 

 efficiency, such that its axis was parallel to that of the collimator 

 and telescope. 



A vertical arc or horizontal spark image was thrown upon the slit and 

 studied visually under numerous and widely different conditions. 

 When a photographic comparison of the two sources was desired, a 

 shutter was used. 



Details of the Investigation. 



Preliminary comparison of spark and arc At the outset an 

 attempt was made to compare the position of the image of a highly 

 disruptive spark with that of the arc. This was soon found to be 

 impossible because of the fact that the lines given by a disruptive 

 spark between terminals of the pure metal were not sufficiently 

 monochromatic. Their images given by the Petitdidier instrument 

 cannot be distinguished from those given by the corresponding region 

 of the spectrum of a Nernst lamp (see Plate 1, 52) and the position of 

 the maximum intensity is a function of the condition of the echelon 

 whether purely of a single or purely of a double order nature at the 

 temperature of the instrument. The only cases in which this method 

 would apply are those in which the spark line is more nearly mono- 

 chromatic and the condition is absolutely that of a single order. 

 Even then the form of the intensity curve for white or not fully mono- 

 chromatic light would have to be known. 



Visual study of arc lines. ^^ As the conditions in the arc and 

 the resulting structure of the lines of the spectrum often change very 

 rapidly, it appeared to be of interest to study these three strong zinc 

 lines visually. A study of this sort was made, an assistant keeping 

 the arc image on the slit and recording the structure of the line as 

 dictated to him. From various sets of observations, many of which 

 are mutually confirmatory, the conclusions given below may be drawn, 



18 These visual observations were made in a wholly unprejudiced state of 

 mind for, although the papers of Nutting and Janicki referred to had been 

 read when they were first published, the details of the same had been quite 

 forgotten by the writer of this paper. 



