296 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



13. In iron, Table VI, we find that, among the lines investigated, the 

 shifts are in general very small, but are nevertheless present. 



Further facts are as follows: 



14. A plate on which three exposures were made in the usual manner, 

 the first and third being superposed, and the source of light being the 

 titanium arc, shows that no shift is introduced by throwing a diffuse image 

 on the slit, destroying the focus by moving the arc two inches nearer the 



TABLE VI. Iron* 

 Shift of Spark line toward Red in Tenth-meters. 



lens. This proves that even if either arc or spark were out of focus 

 to a small extent the shift would not be altered. 



15. A plate exposed to the carbon and graphite arcs shows that no 

 shift occurs between lines which are due to the same impurities in both. 

 The currents used were 33 and 75 amperes respectively. The graphite 

 arc is produced with considerable difficulty, a large current bei»g required, 



* The conditions under which the iron plates were obtained are as follows : Six 

 plates taken at various times ; average spark exposure, 12 minutes ; average current, 

 power, and voltage, 41.3 amperes, 420 watts, 11.5 volts; capacity, 0.0226 microfa- 

 rads ; length of primary gap, 9 mm. ; no secondary gap and no fan ; region of spark 

 image used, that near terminal. The above values of the shift are means of 

 measurements made on from two to six plates. 



