THE EFFECTS OF PRESSURE UPON LINES IN THE SPECTRUM OF 
IRON. 
BY FRANK F. ALMY. 
Humphreys and MohieF, in 1896-7, made an extended study of the 
effect of pressure upon the wave frequencies of the lines of the emission 
spectra of a larger number of the metalic elements. The source used was 
an enclosed electric arc about which the air pressure could be varied, the 
pressure being carried up to fourteen atmospheres. The results of their 
experiments indicate that all isolated lines shift toward the red end of 
the spectrum with increased pressure, the shift being proportioned to the 
wave length for lines of a given series, but the factor of proportionality 
being different for lines of different series. 
In the winter of 1901-2 at the suggestion of Professor Michelson, I 
undertook the measurement of the pressure shift of some of the lines 
of the spectrum of iron by a method quite similar to that employed by 
Humphreys and Mohler, but with essential differences. The source of 
light employed was a self induction spark produced by rotating a sprocket- 
wheel by a steel spring. The spark gap was contained within a cast 
brass vessel with properly packed shaft for driving the sprocket wheel 
and insulated leads, and a plate glass wundow. An image of the spark 
was formed upon the slit of a grating spectrometer. The spectrum was 
formed by a 5-inch Rowland concave grating of 21 feet radius ruled 
20,000 lines to the inch. My study was confined to the second order spec- 
trum in which the dispersion was such that 1 Angstrom unit was approxi- 
mately 1 mm. in the spectrum; and to that part of the spectrum of iron 
lying between wave lengths 4014 and 4528. The photographs were made 
by the ordinary method of three exposures to check any accidental dis- 
placement of the plate. 
The pressure shifts found for the lines in this region conformed very 
closely to those published by Humphreys. Of the iron lines in this reg- 
ion certain of the broader, denser lines are reversed at their centers in 
the spectra under pressure. This is particularly evident in the lines 
4045, 4063, 4071, 4271, 4308, 4325, 4383 and 4404. It may be noted that 
these reversals agree with those noted by Professor Hale^ when the spark 
gap was under pressure in a liquid, and later observed by Hale and KenU. 
After a somewhat extended study of this region of the spectrum of 
iron by means of the grating spectrometer, I attempted to apply the 
echelon spectrometer to the problem of pressure shift. 
The same source was employed as before, the light first passing 
through a prism spectrometer using two 60° carbon bisulphide prisms 
in succession, which gave sufficient dispersion so that lines separated by 1 
( 231 ) 
