FEBRUARY 11, 1897 | 
NATURE 
345 
The insulated metal was afterwards charged positively, 
and then negatively. The rate of discharge was ob- 
served till the ultra-violet-light-zero was reached. 
With polished zinc as the insulated metal the following 
results were obtained. 
The insulation was first tested. When no ultra-violet 
light was used it was found that the electrometer reading 
remained the same whether the two pairs of quadrants 
were in metallic connection or not. With the ultra- 
violet light shining the reading with the quadrants in 
metallic connection was the same as before, the readings 
with the quadrants disconnected were :— 
Januaey 14th, 3h. 41m. p.m. 
25 sc. divs. from metallic zero after rs ees, 
= 45 ” ” ” 3° ” 
— 59 33 » of 45 
— 67 D 35 aa I min. 
= 80 ” ” ” Tay > 
- 89 » » » 25, 
- 99 ” ” 2 eae) 
= TOT ” 4 55 
"Afterwards ste: ady. 
[Sensibility of electrometer, 140 sc. divs. per volt. ] 
The difference thus found, between the metallic zero 
and the ultra-violet-light-zero, is —1or1 or —o'72 of a 
volt. 
3h. 47m, Zine charged positively to 
219 scale divisions from the metallic 
zero, 
Reading from metallic zero with ultra- 
violet light shining. — 
Time. 
| first 4 cms. 
| as not to illuminate either directly. 
tell us what would happen if an insulated metal, shone on 
by ultra-violet light, were surrounded by a metal of 
precisely the same quality of surface connected to 
sheaths. 
So far we have mentioned only experiments in which 
the rays, whether Roéntgen or ultra-violet, fell perpen- 
dicularly on the insulated metal. We have also made 
some experiments with the rays going parallel to the 
metal surfaces. 
For this purpose a cardboard box 46 cms. long, 19 
cms. square (see Diagram 3), lined, in the first instance, 
with tinfoil, connected to sheaths, was used. Inside this 
box an insulated disc of oxidised copper of 10 cms. 
diameter was supported in such a way as to allow of its 
being fixed at different distances from the tinfoil-coated 
end-wall of the box facing it. 
The distance between the disc and the tinfoil was at 
The arc lamp was distant about 20 cms. 
from the box. The light from it shone through a slit in 
| the tinfoil covering the side of the box perpendicular to 
the surface of the oxidised copper. The slit was 4 cms. 
long, 1 cm. broad. Its length was first placed parallel 
to the copper surface, so that the light admitted by it 
shone in the space between the two metals in such a way 
It was found (1) 
that the ultra-violet-light-zero did not deviate from the 
Wecteo meter. 
+12 after 15 secs. 
+ 64. 0m “Fi » 39 55 
+ 23 ee oD » 45 » 
— Igstar ees ” I min 
= Obs ie es ” 1} ” 
= 7 Ogu =e as 
= O35) ues ait Pe 25) >> 
—100 ... ae om 3e 
— 103 4 55 
is onch ere 
Afterwards steady. 
3h. 55m. Zine charged negatively to 
238 scale divisions from metallic zero :— 
—177 sc. divs. from metallic zero after 15 secs. 
— 149 ” ” ” 30 ” 
— 132 ” ” ” 45 2 
—124 rr) an) 3g I min. 
— 113 ” ” ” 2 55 
—iIil ” ” ” 3 
Afterwards steady. 
The following table shows the steady potential 
differences in the electrometer due to the conductive 
effect of ultra-violet light in our apparatus between the 
brass wire gauze and plates of various other metals. 
Insulated metal :— 
Polished zinc —o'75 of a volt. 
Polished aluminium —0'66 rf 
German silver ... —O 19 a 
Gilded brass +0°04 ny 
Polished copper preva: oe 
Oxidised copper 4 TOS ss 
The copper was oxidised by being held in a Bunsen 
flame. 
In the case of polished zinc, polished aluminium, 
polished copper, and oxidised copper, both positive and | 
negative charges were discharged at the same rate, if we 
reckon the charge of the insulated metal from its ultra-_ 
The rates of reaching the ultra-violet- | 
| violet-light-zero was found to depend on the distance 
violet-light-zero. 
light-zero were not observed for gilded brass and German 
silv Sie, 
It must again be noticed that our experiments do not | 
NO. 1424, VOL. 55 | 
DIAGRAM 3. 
metallic zero when the sheet of light passed between the 
two metals ; (2) that a negative charge given to the in- 
sulated oxidised copper was not discharged ; and (3) 
that a positive charge was removed very slowly—about 
four scale divisions per minute from a charge of 197 
scale divisions from the metallic zero. 
When the length of the slit was placed perpendicular 
to the surface, so that a small portion of both metals, as 
well as the intervening air, was illuminated, it was found 
that the reading deviated about +1 scale division per 
minute from the metallic zero. The oxidised copper 
was charged positively ; and negatively. Discharge took 
place at about four scale divisions per minute, from a 
charge of +202 scale divisions ; and three scale divisions 
per minute from a charge of —246 scale divisions: the 
charge reckoned from the metallic zero in each case. 
The slit was then so arranged as to allow the light to 
shine on the oxidised copper alone. In this case the 
deflection went towards an _ ultra-violet-light-zero at 
about +6 sc. divs. per minute; and both positive and 
negative charges were discharged, the negative much 
more quickly than the positive. 
The ultra-violet light was now shone between the 
oxidised copper and the disinsulated tinfoil wall opposite 
to it, parallel to their surfaces so as to illuminate both. 
The difference between the metallic zero and the ultra- 
between the two surfaces. This will be seen from the 
following table :— 
