410 TRANSACTIONS OF SECTION A. 
3. The Effect on the Persistence of Vision of Fatiguing the Hye with 
Red, Orange, and Yellow. By Professor Frank AuuENn, Ph.D. 
In order to determine the persistence of vision a sectored disc is rotated 
in front of the slit of a spectrometer at such a speed that the flickering of 
the colour is made to just disappear. This critical speed is very accurately 
measured electrically by means of a chronograph. A persistency curve can 
then be obtained which is shaped like a parabola with its apex at the point 
corresponding to the brightest part of the spectrum. If the eye is constantly 
fatigued with some colour and the persistency now measured, the two per- 
sistency curves when plotted together show some peculiar and characteristic 
differences. When the eye is fatigued with red light of wave-lengths 680 
and 670, the fatigue curve differs from the normal only in the part 
corresponding to red. When the eye is fatigued with green, the two curves 
differ in the green alone. But when the fatiguing colour is any part of 
the spectrum between wave-lengths 577 and 650, the curves differ in both 
red and green. When, however, the wave-length 660 is used as the fatiguing 
colour, the two curves coincide completely. This means that the funda- 
mental red sensation is at least beyond wave-length 660, and that yellow 
and orange cannot be simple primary sensations. 
4. A New Method of Measuring the Luminosity of the Spectrum. 
By Professor Frank ALLEN, Ph.D. 
The principle upon which this method depends is that the persistence 
of a colour sensation on the retina is a function of the luminosity of the 
colour only. 
Two Nicol prisms are mounted in front of the spectrometer slit. It is 
so arranged that in the eyepiece one sees two small adjoining patches of 
light, one white, the other all colours of the spectrum in succession. The 
white light is of such low intensity that its persistence is the same as that 
of very weak violet light. By rotating the polariser the intensity of the 
spectrum colours is reduced to that of the standard white. The intensity 
of the light going through the Nicols being proportional to the square of 
the cosine of the angle between their principal planes enubles the luminosity 
to be determined. 
A sectored disc rotating in front of the slit interrupts both white and 
coloured lights at the same time, thus enabling the persistence of the flashes 
of light to be measured. 
5. The Effect of Low Temperatures on Fluorescence Spectra. 
By Professor Epwarp L. Nicnous and Ernest Mernritv. 
This is a quantitative spectrophotometric study, by methods previously 
employed by the authors and described in various papers, of the spectra 
of certain fluorescent substances at temperatures between +20° C. and 
—185° C. 
The bodies subjected to measurement were :— 
(1) A specimen of natural willemite in powdered form; 
(2) Commercial anthracene showing the green fluorescence bands of the 
impure preparation ; 
(3) An alcoholic solution of fluorescein ; 
(4) An alcoholic solution of resorufin. 
Excitation was produced by means of a quartz-mercury lamp any portion 
of the spectrum of which, dispersed through quartz, could be focussed on 
the surface of the fluorescent hody. The latter was placed in an enclosed 
