100 GOODSPEED—FIELD SURROUNDING CROOKES TUBE. [May 15. 
influence, for the density of the exposure was found to be the same 
over the area within as under the edge of the tube, z.e., nearly zero. 
It does not seem possible that this effect could result entirely 
from the action of the Sagnac rays, since little if any of the area at 
the base of the brass cylinders can be reached by a straight line 
from any particle of matter traversed by the direct X-rays. It can 
be explained as a tertiary effect, produced by the air or wood just 
over the top, which had received its energy from the secondary 
emanations of other bodies in the direct path of the X-rays, or 
possibly the secondary or Sagnac rays may be of the nature of dark 
phosphorescence, 7.¢., lasting for a time after the cause has ceased. 
Reasons for favoring the latter view appear as a conclusion to this 
paper. In this case the diffusion of the air in the room would cause 
the whole space to be uniformly active. 
The arrangement just described suggested some easy tests on 
reflecting or diffusing power of different surfaces, as Sagnac had 
made in a different way in his investigations. In a brass tube 
similar to the one used above, at two points 90° apart, windows 
were cut 1 centimetre wide and 4.5 centimetres high. This tube 
was capped with zinc or lead, so that nothing could enter except 
through the windows. It was placed on the plate and a polished 
zinc block arranged opposite one window. Fig. 9 shows the result, 
all other conditions being as before. The exposure was twenty 
minutes. The streak entering the window opposite the zinc is 
unmistakable, and the diffused ‘‘radious’’ state of the whole 
enclosed space is demonstrated by noting the line of contact of the 
tube. A little brush in at the other window, too, is clearly dis- 
tinguishable though faint. It seemed most desirable now, if pos- 
sible, to make this phenomenon optically visible, and with this 
in view the following arrangement was set up: 
Instead of the smaller lead block used with the radiographic 
plates, sheets aggregating one centimetre in thickness and a little 
larger than a 7xg screen were placed on the box. On this a barium 
platinum cyanide screen was placed, face up, but covered with a 
piece of pasteboard. In this cover a circular hole was cut just the 
outside diameter of the window tube described above, through which 
the latter was placed, resting on the fluorescent screen. Its length 
was doubled by placing an extension on top. This was found by 
experiment effectually to exclude all noticeable influence except 
that through the windows. 
