PHYSICS: NICHOLS AND HOWES 
307 
It is understood that in these specifications and elsewhere, unless otherwise 
stated, the curve of decay is plotted with time from the close of excitation as 
abscissae and the reciprocal of the square root of intensity (7~ 2 ) as ordinates. 
A linear process is any straight portion of the graph; as 1, 2 or 3 in figures 1 
and 2. The inference is that the more or less abrupt changes of slope are re- 
lated to and indicative of real changes in the processes by which the emission 
of light from the phosphorescing body is being maintained. 
A distinguishing criterion of type is found in the sign of the intercepts 
of the various processes at the origin of time. In type 1 the intercepts are all 
positive; in type 2 the intercepts of process 2 and 3 are negative. 
I * 
CALCITE. 
ULTRA-VIOLET EXCITATIOM. 
SEC. 
FIG. 3 
FIG. 4 
The passage from one process to the next is presumably never actually dis- 
continuous, as in the above diagrams, but it is sometimes very abrupt as in the 
curves of figures 3 and 4 which are from our recent measurements of the phos- 
phorescence of calcite. Sometimes the transition from slope to slope is very 
gradual; so that the first and second processes tend to merge into a curved 
line. We have shown, 6 in the case of the uranyl salts that the location of 
the knees depends on the intensity of the excitation and that distributed knees 
— so to speak — occur when the exciting light penetrates the crystal and being 
absorbed produces luminescence of decreasing intensity in successive layers; 
also that where the conditions are such that excitation is confined chiefly to 
the surface, the knees are well defined. 
Similar modifications of the curve occur in kathodo-excitation, and distrib- 
uted knees are then probably due to a kathodic discharge containing par- 
ticles of varying velocity. 
