4 
; 
f. 
3 
O. N. Rood on a Secondary Spectrum. 179 
indifferent. To determine the amount of the difference due to 
an exchange of the position of the prism and grating, I made a 
few experiments which may now be detailed. : 
and the secondary spectrum A was obtained. The whole 
apparatus was then re-arranged, and the same experiment re- 
peated, with the result given in B. 
Na Th Li HB Hg* 
A Pie | 8 Yo yee ae 
Wy ee ae Li HB Hig* 
B lo 4 es | 88°] 748 | ITT | 
_ On comparing these secondary spectra with those obtained 
in the 3d case, it will be seen that the differences are not larger 
taan the necessary errors of observation, except 1n the distance 
between Hf and Hg*, where the difference is about twice as 
great as it otherwise ought to be. A second perfectly similar 
experiment was now made with flint glass prism and grating 
No. 2, with the result given in C. 
Th He Na 3 Ha Li Hg! 
gare ie Bo: ak a Se o-95 | 1°75 | 11 | 285 | 
On comparing this secondary spectrum with that obtained in 
errors, and that hence the above construction, practically, ap- 
plies to the secondary spectrum furnished by #vo prisms of dif- 
ferent materials, 
