154 THE ROYAL SOCIETY OF CANADA 



Working with pure hydrogen excited also by electromagnetic induc- 

 tion, but with spark gap as long as 2 inches, and with pressures some- 

 what higher than those used by the writer, Masson obtained a rose 

 ring showing both line and secondary spectrum, together with an 

 inner blue zone whose spectrum showed nothing but lines. 



It would seem, therefore, that in pure hydrogen four distinctly 

 different coloured discharges are possible: (1) whitish, in which the 

 Balmer lines are almost absent; (2) pink, in which both line and 

 secondary spectra are strongly developed ; (3) red, in which Ha and 

 H^ are relatively very strong and at least a portion of the secondary 

 is absent; (4) blue, in which, according to Masson, the secondary is 

 entirely absent and, moreover, iJ/3 is stronger than Ha. Concerning 

 this last appearance the writer will make further observations to see 

 if Masson 's results can be confirmed. 



Discussion 



The interpretation of these results is not easy. It is generally 

 assumed that the Balmer lines have their origin in the atom, the 

 secondary spectrum in the molecule. But the secondary spectrum is 

 very complex. Fulcher,^ for example, has shown that in the region 

 approximately X 6500-X 5380 the lines may be divided into two 

 groups, one of which is weak at feeble excitation, the other of equal 

 intensity at feeble and at strong excitation. Merton,^ too, has 

 divided the many lined spectrum into three groups, the first of which 

 is unchanged, the second enhanced, the third Aveakened, by the 

 addition of helium to hydrogen. The writer's results indicate that 

 in a region where the Balmer lines are relatively strong (a condition 

 which is only obtained with strong excitation), a portion of the many 

 lined spectrum, as well as the continuous background, is feeble or 

 absent altogether, while another group of these (between X 4316 and 

 X 4136) is strongly developed. If, now, we make the assumption to 

 which little objection can be made, that the atom is the origin of the 

 Balmer lines, this suggests that at least some of the lines of the 

 secondary also may be associated with the atom. On the other hand, 

 it is well to remember that in strongly excited gas we may have not 

 only neutral atoms and molecules, but ionized molecules. Moreover, 

 it has been shown by Franck^ that the work of dissociation plus the 

 ionization of an atom is less than the work required to ionize a mole- 

 cule. Accordingly ionized molecules do not appear until the excita- 



6G. S. Fulcher, Astrophy. Jour., Vol. 37, p. 60, 1913. 

 «Merton, Proc. Roy. Soc. A, 679, p. 382, Jan. 2, 1920. 

 'J. Franck, Phys. Zeit., No. 16, S. 466, 1921. 



