Sia es | >= 
sy, June 2, an NATURE 
+. 
ee 
=. == 

743 

This sequence of changes takes place only when the 
circuit can oscillate. To follow the complete sequence 
it is necessary either to shunt the galvanometer 
heavily or to-substitute for it a milliamperemeter. 
A set of observations so obtained is given in Fig. 2, 
which represents the apparatus in the best adjustment 
for our purpose. To obtain this linear form of curve 
300 

(Shunt = 600.) 
Ga lvanometer 
anode current for no oscillations” 
Cms. “OR 
Fic. 2. 
“Ol 03 
the relative positions (“ coupling ’’) of the coils must 
be altered, and the most suitable plate voltages (V) 
ascertained by trial. With some valves it may be 
necessary to apply a negative potential to the grid 
between the points g, gs, Fig. 1. I have used 
several common makes of ‘“‘ R Type,’ hard valves, 
and have never found any difficulty in obtaining the 
condition shown. 
. The function of the “ zero-shunt ”’ is to by-pass an 
amount E/R of the anode current 7, so that, if R is 
large compared with the resistance of the galvano- 
meter, the current through the latter is approxi- 
mately 7 -E/R. When E and R are chosen so as to 
make this difference small, a sensitive galvanometer 
can be employed, unshunted, which will then give 
“at deflexions, when the plates of the condenser 
suffer minute displacements. In view of the linear 
form of the curve (Fig. 2), it will be clear that the 
lvanometer reading is proportional to the plate 
isplacement. Also, calibration is readily obtainable 
by shunting the galvanometer, say ten times, and 
then observing the deflexion obtained when the 
micrometer screw is turned through, say 1/1000 cm. 
A resistance y up to 1000 ohms may be introduced 
into the oscillation circuit to reduce the sensitivity 
and widen the range of the plate movement. This 
Tesistance also renders it easier to obtain the linear 
adjustment. The large black dots in Fig. 1 represent 
the terminals on the case of the instrument; the 
batteries, galvanometer, and condenser plates are 
external and are connected to these terminals. The 
actual resistances, coils, etc., are mounted beneath 
the ebonite top of a small box, about one foot square 
and a few inches deep. Rigid connexions are em- 
ployed to eliminate vibrational effects. 
In the use of the apparatus for recording small 
displacements, movements, etc., one of the condenser 
NO. 2796, VOL. 111] 
plates is caused to partake of the movement to be 
measured by direct attachment, if possible, to the 
moving member. The other plate may be mounted, 
as already described, on a micrometer screw device 
to facilitate calibration. For steady working, at all 
times great care must be taken to employ batteries 
that are in perfect condition, and have an adequate 
current capacity. It is advisable to use cells of the 
Same type for E as for V. Temperature changes 
must naturally be avoided in view of expansion 
| and other effects. For ‘‘super’’ sensitivities (above 
10-7 cm.) screening and other precautions become 
necessary. Joun J. Dow Linc. 
University College, Dublin, May 7. 

A Permanent Image on Clear Glass. 
THE interesting observation described by Mr. Eric 
Robinson in Nature of April 28, p. 569, and com- 
mented upon in the same issue by Dr. J. W. French, 
is an excellent example of the ease with which the 
surface of glass may suffer modification and retain 
it over a long period of time. The present writer 
has studied a number of phenomena connected with 
‘breath figures,’’ and an account of the work will 
be found in the Philosophical Magazine for October 
last. : 
If the tip of a small blowpipe-flame is drawn 
rapidly across a sheet of glass it can be shown in 
various ways that the surface of the glass along the 
flame-track has been considerably modified. Flames 
of coal-gas, carbon monoxide, and hydrogen produce 
identical results. When moisture from breath con- 
denses on the glass it is in the form of a misty deposit 
of minute hemispherical droplets, except along the 
flame-track, where it collects as a continuous trans- 
parent film. The contrast between the two types 
of condensation is most marked and constitutes a 
“breath figure.’’ These flame-tracks are revealed 
when silver is chemically deposited upon the glass 
and they can also be traced by the greatly increased 
friction which manifests itself when a chemically 
cleaned watch-glass, which is being dragged across 
the plate, encounters one of the tracks. The in- 
sulation of the glass surface is also less along a flame- 
track than it is on those parts which have not been 
exposed to the action of the flame. = 
It is not possible in the space available to give 
the evidence in favour of the conclusion reached by 
me that at least two causes operate in producing 
the modification of the glass surface which leads to 
a ‘‘ breath figure.’’ One of these is that the flame 
removes the extremely thin film of contamination 
which certainly covers all glass which has not been 
subjected to a rigorous chemical cleansing process, 
and the other is probably a physical change in the 
surface of the glass itself. The latter is very persistent 
and can be detected for many months after the 
passage of the flame across the glass. I am inclined 
to attribute Mr. Robinson’s effect to a physical change 
in the glass surface. Is it not possible that the 
gelatin of a photographic print which has been 
squeegeed upon glass may, when dry, exercise a 
considerable force on the surface in contact with it 
and that this force may have different local values 
depending upon the density of the photographic 
image? Such local differences in tension may impress 
upon the glass corresponding differences in surface 
structure which would then be capable of detection 
as a ‘“ breath figure ’’ or by deposition of silver. 
T. J. Baker. 
King Edward’s School, Birmingham, 
May 7. 
