32 B. E. Moore 
recorded p-components have very uniformly double, the intensity 
of the s-components. 
VIII. Comparison OF SUBSTANCES 
It is desirable to compare the substances discussed above with 
lines whose separation gives series. The two yttrium lines, 4395.91 
and 3195.80, which are accurate duplicates such as one expects 
in series, are like six of the nine components of mercury, 5461 of 
the second adjacent series. The quadruplet 4167.65 is like four 
components of the barium sextet 5854. The quadruplet 3982.75 
is like four components of the thirteen-component mercury tine 
3063.5. The remaining components in these three lines are not 
absent because of weakness, since these lines are not weak lines. 
The two eleven-component zircon lines have two pairs of com- 
ponents similar to two pairs of components in the nine-component 
mercury line 5461. They have two other pairs of components 
represented in the barium principal series greater wave-length 
line 4934. Their one remaining pair of components is found in 
harium principal series smaller wave-length 4554. That is, the 
components of these two lines are represented in three series 
types. There are other lines both in yttrium and in zircon whose 
components can be selected by using more than one line of barium 
and mercury. Such a comparison points to a difference in the 
character of the lines rather than to an identity. From a glance 
at the tables it is seen that the components of lines are multiples 
of intervals. Further, that the lines under comparison are like- 
wise multiples of the same intervals, but that the multiples are 
only partially identical when they have the same interval and the 
components have the same multiple proportions. One may further 
add, the components should have the same relative intensity. This 
is the condition fulfilled in the principal and subordinate series. 
The quadruplets of the principal series are four- and two-fold the 
interval +-.37 (==Ad/X*), for field strength 24400, and the sex- 
tuplets five-, three-, and one-fold the same values. 
While I was engaged in making these comparisons an inyesti- 
gation bearing directly upon this point was published by Professor 
us) 
; fe a ee, el ee a oe 
a 7 CSP) eae 
4 6 iy 4 
i 
{ ; 
‘ 
