868 Mr. Gr. A. Hemsalech and the donate de Gramont on 



line. Further, the lines of the Rydberg series form a pro- 

 minent feature of the spectrum. It is suggested that the 

 emission of A 4481 and the more pronounced appearance of 

 the lines of the Rydberg series are partly caused by the 

 chemical reactions involved in the formation of magnesium 

 nitride. § 5. 



8. The spark line X 4481 is brought out prominently in an 

 arc burning under liquid air and its emission takes place 

 evenly along the whole path of the flash. Under similar 

 conditions the spark lines of zinc show strongly only during 

 the first phase. § 6. 



9. Experiments and observations are described which seem 

 to indicate that the broadening of the hydrogen lines H a and 

 H/3 observed during the early stages of an arc flash is caused 

 by the strong electric field set up momentarily when the 

 electrodes separate. § 7. 



10. A determination of the intensity of the electric field 

 set up during the first phase of a water arc made from the 

 amount of displacement of the outer components of H^ has 

 given a value which is in good agreement with that obtained 

 by direct calculation. § 7. 



11. Detailed results are given of observations of the spectra 

 of magnesium, zinc, and cadmium under various arc con- 

 ditions. Reproductions of photographic records secured in 

 the course of this research illustrate the relative behaviour 

 and the characteristic features of various types of lines. 

 §§ 8 and 9. 



12. A discussion of our results with particular reference 

 to the line \ 4481 of magnesium has led us to the conclusion 

 that its emission during the first phase of a liquid or vacuous 

 film arc is connected with the momentary establishment of a 

 strong electric field between the electrodes. The further 

 suggestion is made that the symmetrical broadening of this 

 line and the expansion towards the red of the lines of the 

 Rydberg series, which are observed under these same con- 

 ditions, are manifestations of the Stark effect. § 10. 



§12. Concluding remarks. . :„. 



The fact that spark lines or so-called high temperature 

 lines are brought out prominently in an arc when the vapours 

 of the latter are cooled, and that they disappear, on the other 

 hand, when the temperature is raised, should help to remove 

 the last douhts as to the futility of using these particular 

 spectrum lines for the estimation of the temperature pre- 

 vailing in a source of light. Further, the emission by the 



