474 Prof. R. W. Wood on *7i« 



vapour. In the earlier work upon iodine it was thought 

 that the series excited by the longer o£ the two yellow lines 

 coincided with the series excited by the green line. The 

 higher dispersion used in the present work shows that this 

 is not the case, for although coincidence is so nearly 

 approached in the vicinity of the yellow lines that only 

 the sharpest photographs taken with such a fine slit that 

 their reproduction is impossible, show separation, the larger 

 value of the spacing in the case of the series excited by the 

 yellow line 5790 causes the lines of this series to get so far 

 out-of-step with the lines of the series excited by the green 

 mercury line, that the failure of coincidence is clearly visible 

 in the photographs reproduced, in the red region of the 

 spectrum. (Compare figs. 3 and 4.) 



The transformation of the resonance into the band spectrum 

 is shown in the photograph reproduced in fig. 5, with the 

 neon spectrum superposed. The darker portion of the 

 spectrogram has been intensified with mercury bichloride, to 

 bring out the bands more distinctly. This photograph was 

 taken with a large glass bulb containing helium at 5 mm. 

 in addition to the iodine. The behaviour of the other gases 

 of the helium group has been investigated in this respect 

 and will be discussed presently. 



The wave-lengths of the lines in the various series have 

 been determined with great care, from a large number of 

 photographs. Short exposures were made with a very fine 

 slit for the strong lines, since the long exposures which were 

 necessary to bring out the faint lines resulted in so much 

 broadening of the stronger lines that their exact measurement 

 was impossible. 



The wave-lengths of the lines are given in the following 

 Table. In column A we have the series excited by the 

 green line 546, in column B the series excited by the yellow 

 line 5769*5, and in column C that resulting from the yellow 

 line 5790'5. The exciting lines are printed in heavy type. 

 In columns a, b, and c we find the wave-length differences 

 of the lines of the corresponding series. Column d gives 

 the wave-length differences between the members of the 

 15 pairs of lines excited by the two yellow lines (the B and 

 C series). 



As appears from the table the wave-length differences 

 have a tendency to become larger as we approach the red 

 end of each of the series. There is also a certain amount of 

 oscillation, thai is, the differences do not increase steadily 

 and uniformly. I am of the opinion that the wave-lengths 

 can be considered correct in most cases to an Angstrom unit 



