562 
Kaye and Higgins have extended these furnace 
experiments to the case of the alkaline earths 
and a number of metals. When these substances 
were suddenly introduced into the furnace, even 
more remarkable electric currents were recorded, 
amounting at 2500°C. to about 4 amperes per sq. 
cm. with barium oxide, about 2 amperes per sq. cm. 
with boiling tin, and rather less with boiling iron. 
Without a doubt, the emission with baryta is 
one owing its magnitude in great part to chemi- 
cal action; and this view, it may be noted, tallies 
with the opinion now generally held that the 
activity of this substance (and of lime) in the 
Wehnelt kathode, is effected largely through the 
intermediary of the residual or occluded gas, or 
possibly of the platinum kathode. 
These conclusions as to the part played by the 
surrounding gas are on all fours with the results 
of recent work by Fredenhagen and Kistner 
(Phys. Zeit., January 15, 1914), and Hallwachs 
and Wiedmann (Berl. Ber., January, 1914) on the 
photo-electric effect. These observers took steps 
to renew continuously (either by scraping or dis- 
tillation) the surface of a metal (Zn and K respec- 
tively) in a vacuum of exceptional excellence, and 
then found that, if elaborate precautions were 
taken to remove any occluded gas as fast as it 
was released, no electronic emission such as is 
ordinarily produced by ultra-violet light could be 
detected. The inference is that chemical action 
plays a prominent part in photo- Peco ay ihe as 
in many thermionic experiments. 
But ‘there are certain. cases described by 
Richardson and his coadjutors, and more recently 
by Langmuir (Phys. Rev., December, 1913), where 
the temperature factor is apparently competent to 
explain. quantitatively the observed phenomena; 
and the present position appears to be that several 
effects may be concerned in tke generation of 
electricity by hot bodies, viz. 
(1) That due purely to Epa anes: 
(2) That due to chemical reaction; 
ably 
(3) That due to a change of state—volatilisation 
and possibly liquefaction. 
It does not follow that all the various causes 
will conspire to help each other. For example, 
Kaye and Higgins noticed that boiling brass gave 
out at 2500° C. a large emission of positive elec- 
tricity, and it may be that in this case the vapor- 
isation effect was positive and sufficiently great 
to mask completely the negative emission due to 
other causes. 
The part played by volatilisation has not re- 
ceived so much attention as the others, but the 
experiments at the National Physical Laboratory 
offer support in favour of this view, which, after 
all, is but an extension of what is well known in 
cases of bubbling and splashing of liquids at 
moderate temperatures. 
The Directive Volatilisation of Metals.—Evid- 
ence of the volatility of metals at temperatures 
well below their melting points is of long stand- 
ing. A familiar illustration is furnished by the 
blackening of tungsten and carbon filament lamps. 
NO: 2325, OL.«03)| 
and_prob- 
NAT GORE 
[JULY 30, 1914 
| : Ae : 
| Deposits of definite outline can often be detected 
on the bulbs of the lamps, and the blackening 
frequently takes the form of a parallel band of 
deposit which is confined to the glass immedi- 
ately opposite the windings of the filaments. The 
fact seems to point to the projection of particles 
in definite directions from the filament. 
We are led to the consideration of a number of 
researches, which have established the fact that 
the particles which are given off from the surface 
of a metal during volatilisation tend to travel in 
straight lines at right angles to the surface. This 
rectilinear emission appears to have been first — 
noticed by Dunoyer (Comptes rendus, 152, p. 
592) at Paris in 1911. In his experiments a piece 
of sodium metal was placed at the bottom of a 
vessel which was highly exhausted. Above the 
scdium were mounted two parallel screens, in 
each of which was a small hole, the one being 
vertically above the other. The sodium was 
heated to about 400° C., so that the liquid metal 
was vaporising freely, though not actually boil- 
ing. Dunoyer found that if a small obstacle was 
placed above the upper diaphragm a _ shadow 
(with umbra and penumbra) was clearly traced 
out in the deposit of sodium condensed on a 
screen above the upper diaphragm. Evidently the 
metal was propagated in straight lines, and 
Dunoyer looked to the individual molecules as the 
carriers of the metal, these molecules being able 
to maintain a straight course owing to the low 
gas pressure. 
But similar experiments carried out by Reboul 
and de Bollemont (Journ. de Phys., July, 1912) 
cannot be quite so simply explained. These ex- 
perimenters mounted vertically within an electric 
furnace two small sheets of metal facing each 
other and a few millimetres apart. One sheet 
(which was usually of platinum) acted merely as 
a receiving screen, while the other consisted of 
the metal the volatilisation of which was being 
studied. With the furnace below 400° C., no 
results were obtained; but at temperatures be- 
tween 400° C. and goo® C. it was found that if 
the volatilising sheet was of copper or silver, a 
black deposit, which closely followed the shape of 
the emitting metal, was obtained on the screen. 
The extent of the effect increased rapidly with 
the temperature. In air at atmospheric pressure 
the best results were secured at about 1 mm. dis- 
tance, 3 mm. being the greatest distance at which 
definite deposits were secured. 
Fig. 1 shows the cruciform deposit obtained in 
air at atmospheric pressure from a copper sheet 
cut in the form of a small cross. In this example, 
the furnace temperature was 850° C., the range 
1 mm., and the time of exposure 30 secs. 
Reboul and de Bollemont repeated the observa- 
tions under various conditions. In oxygen the 
effect was enhanced; in a vacuum, the deposit 
gained in sharpness of outline. Curiously 
enough, in hydrogen, the edges of the strip 
seemed to be the only active regions, so that the 
deposit merely reproduced the outline of the 
strip. 
‘ 
£ 
4 
