
Jone 9, 1923] 
NATURE 
775 

Denning examines the dates of abundant showers of 
Perseids and finds that a large proportion of them 
can be satisfied by a period of 11°75 years. This 
period would give an abundant shower for the year 
1592, and Mr. Denning has included that year, 
ty on Mr. Beveridge’s authority, in the list 
of years of observed maxima. A list of Perseid 
showers, or at least of showers at the time of year 
when Perseids are expected, is to be found in Arago, 
“ Astronomie populaire,”’ iv. p. 296-8, and is extracted 
from Biot’s ‘ Catalogue général des étoiles filantes 
. .. observés en Chine,” published in Mémoires 
présentés a l’ Academie, sciences math. et phys., x., 
129, etc. To these Chinese observations it is possible 
to add one from Matthew Paris of the date 1243 July 
26. The complete series then becomes: 714 July 15, 
784 July 10, 830 July 22, 833 July 23, 835 July 22, 
841 July 21, 865 Aug. 1, 924 July 21-23, 925 July 
22, 23, 926 July 22, 933 July 20-25, 1243 July 26, 
1451 July 27, all Julian dates. 
analysis of the dates of the medieval observa- 
tions shows that the date of maximum intensity of 
the Perseid shower has not shifted its position in 
the sidereal year since the year 830 at latest. The 
date corresponds to a solar longitude of 138° reduced 
to the equinox of r900. A few of the showers recorded 
in history fall a little before or after that date, and in 
two instances (784 and 865) the difference is as much 
as ten days on either side of the normal maximum. 
In 1592, on the other hand, the recorded shower falls 
nineteen days earlier than the normal maximum, and 
this raises a doubt whether it was really a Perseid 
shower at all or some otherwise unknown shower 
which apes to fall in a year when an abundant 
_Perseid shower was due. J. K. ForHerincHAM. 
University Observatory, Oxford, 
May 2r. 

The Measurement of Overvoltage. 
In general, the term overvoltage refers to the 
- difference between the potential required to discharge 
an ion at a particular electrode and the calculated 
reversible value, in the same electrolyte. Strictly 
aged therefore, overvoltage only exists while 
t 
e current is flowing, and hence measurements. 
should be made under these conditions. Some 
workers, however, state that the “‘ transfer resistance ”’ 
of a gas film at the electrode causes the measured 
potential to be in excess of the true value; con- 
sequently, an alternative method has been adopted 
in which a rotating commutator rapidly interrupts 
the polarising current and connects the experimental 
electrode with the potentiometer system. In this 
way disturbing influences due to transfer resistance 
are said to be eliminated, since the potential of the 
electrode is measured when the current is not flowing. 
This method gives lower results than the direct method 
for the following reasons: (1) When the polarising 
circuit is broken, an extremely rapid fall of potential 
occurs, which is appreciable even in the small interval 
that elapses between the periods when current flows ; 
(2) since the current only flows intermittently through 
the experimental cell, current density and time effects 
are not comparable with those obtained when the 
current flows continuously ; (3) the continual make 
and break of the circuit by the commutator sets up 
alternating induced currents, and it is well known 
that electrical discharges of such a nature tend to 
lower the potential of a polarised electrode. 
In some recent work, hitherto unpublished, the 
magnitude of the effects due to these induced currents 
has been investigated. The lowering of potential 
NO. 2797, VOL. III] 
was found to depend upon the particular electrode 
examined, and was usually of the order of 0-3 volt, 
and in some cases as much as 0-5 volt. The value 
of the induced current, and consequently its effect 
on the potential, will depend on the frequency of the 
intermittent current, and upon the resistance and 
self-inductance of the circuit; but it seems fairly 
certain that the lowering of potential, due to induced 
currents in the commutator method for measuring 
overvoltage, is considerable. 
In order to eliminate as many sources of error as 
possible, the following method for the measurement 
of overvoltage is being tested. The commutator 
method is being used, but a “‘ choking coil’ of high 
self-inductance is placed in the circuit in order to 
reduce the induced current to a negligible amount. 
Further, instead of the polarising and potentiometer 
circuits being made for equal intervals of time, the 
latter will only be complete for about 10° in each 
revolution. Thus for about 97 per cent. of the time 
the polarising current will flow through the cell, and, 
if the commutator revolves 3000 times per minute, 
only 0-0007 seconds will elapse between opening and 
closing the current circuit. To ascertain the magni- 
tude of the fall of potential during this period, further 
experiments will be made, either by varying the 
speed of the commutator, or by increasing the period 
per revolution in which the experimental electrode 
is connected with the potentiometer system. By 
extrapolation, it should be possible to determine the 
potential of the electrode at the instant of breaking 
the current, and the results compared with those 
obtained while current is still flowing, in order to 
determine the effect of the so-called ‘“ transfer 
resistance.”’ S. GLASSTONE. 
University College, Exeter, 
May 22. 

A New Phototropic Compound of Mercury. 
In an attempt to prepare the phototropic com- 
pound, dimercuric-diiodo-disulphide, described by 
Dr. Ray (Jour. Chem. Soc. 1917, T, 101-109), we 
accidentally “discovered a new phototropic mercury 
compound of the formula Hef through a mistake 
\cns 
of the laboratory attendant in supplying us with 
potassium nitrate in the place of the nitrite. The 
compound is prepared by the interaction of a 
mercuric salt with ammonium sulphocyanide and 
thio-urea in a solution of acetic acid in the presence 
of an oxidising agent. The compound is also pre- 
pared by the action of hydrogen sulphide on mercuric 
sulphocyanide. This gives us a clue to the constitu- 
tion of the yellow mercuric compound. 
The compound is very phototropic, inasmuch as it 
is effected by strong sunlight in less than 1/6oth of a 
second and by diffused daylight in a few seconds. 
It appears, therefore, to be the most phototropic 
compound as yet known. In researches on this 
compound we have prepared a red variety of mercuric 
sulphide by precipitation methods. Again, by the 
decomposition of the yellow mercuric compound, we 
have prepared a yellow variety of mercuric sulphide 
which shows interesting thermotropic properties. 
By the action of free iodine on the new phototropic 
compound, an iodine compound of mercury which is 
also phototropic has been prepared. Further work 
is in progress. Y. VENKATARAMAIAH, 
Bu. S. V. Racuava Rao. 
Research Laboratories, Maharaja’s College, 
Vizianagram, May 9. 
