42 B. E. Moore 
a large range. They do not collect around a “normal separation” 
or multiples of aliquot parts of a normal separation. 
tr. Lines with no components (unseparated) in zirconium show 
no series. Most of these lines are probably unanalyzed types. 
12. There are a great many lines which may be associated in 
pairs. Such lines are comparatively near each cther on the scale 
of vibrations. The pairs may have the same separation or not, 
and have the same or different number of components. In sub- 
stances which have yielded series, such pairs are frequently con- 
spicuous. These pairs repeat themselves in other parts of the spec- 
trum with considerable uniformity of separation. In osmium ren- 
etitions are not present, and in vttrium and zircon only apparently 
so since the separations of the prospective pairs are very irregular. 
X. GENERAL CONCLUSION 
There is a general dissimilarity between the lines of yttrium and 
zircon, and between these lines and the lines of all substances 
which have yielded series. But all substances have common fun- 
damental intervals of small magnitude, and few in number, inti- 
mately connected with a “normal separation.”’ 
In conclusion I wish to thank Professor Voigt of the University 
of Gottingen, Germany, who suggested this investigation and 
kindly placed the resources of the Institute at my disposal, for his 
friendly assistance and enthusiastic encouragement. Likewise, I 
acknowledge my obligations to Professor Runge of the same Uni- 
versity, whose invaluable experience in spectral work was cour- 
teously given me. 
42 
