640 
PHYSICS: F. C. HOYT 
Proc. N. a. S. 
voltmeter was calibrated by measuring the current through a megohm 
resistance in parallel with it, and also checked by measurement of the 
high frequency limit of the spectrum. 
The spectroscopic arrangement consisted of a large calcite crystal with 
a face 5V2 X 8 cm., with a wide slit (1 cm.) placed close to the crystal 
and the plate about 70 cm. from the crystal. The anticathode was about 
20 cm. from the crystal. With this simple setup a greater resolving power 
and dispersion was obtained than with the mica spectrograph used by 
Prof. Webster. Irregularities in the crystal gave little trouble. To 
exclude stray radiation the plate holder was placed in a lead box extending 
nearly to the crystal. In this way a considerable region of the spectrum 
can be mapped at once without moving the crystal; the range from 0.900 
A to 1.24 A in the platinum spectrum, for instance, being spread ov^er 
about 4 cm. 
No attempt was made to measure wave-lengths directly, but for new 
lines the values could be found to about 0.3% by intperolation on the plate. 
Tubes with platinum and tungsten anticathodes and thin glass windows 
were kindly furnished by the General Electric Co. The former contained 
no iridium, but high voltage pictures showed a faint tungsten spectrum. 
L-Series of Platinum and Tungsten. — The platinum and tungsten 
L-series, containing all lines observed with certainty here, is shown in 
Fig. 1. Reproductions of photographs of the region from 74 to are 
shown in Fig. 2, a and h. The tungsten spectrum has received special 
attention, and Dershem,^ Overn,^ de Broglie^ and Siegbahn^ have reported 
a large number of faint lines. All of those recorded by Siegbahn have 
been found, but two of the faint ones recorded by Overn only have not 
been observed. 
In platinum, besides the 13 lines that have been followed through 
from element to element, we find two more lines. and 75, of which 
the wave-lengths as interpolated are 1.080 and 0.988, respectively. jSe 
has been observed by Webster, while 75 has apparently escaped attention 
so far. That these two lines are really part of the platinum L-series is 
supported by the fact that lines occur in corresponding positions in tung- 
sten. Thus the wave-length interval from 75 to 71, in tungsten is approxi- 
mately the same as in platinum, and the same is true of the interval from 
/36 to 185. 
There seems to be a real difference between the tungsten and platinum 
spectra as observed here. With pictures taken under similar conditions 
in the two elements, the relative intensity of the lines is not always the 
same, and some lines occur in tungsten which are apparently absent in 
platinum. Thus ^i, ^s, and (89 and 77 which appear in tungsten are not 
found at all in platinum. For intensity differences, (3^ is much stronger 
than /Se in platinum, but they are of about the same intensity in tungsten. 
There is also a change with 72, which is stronger than 73 in platinum, but 
