576 
NATURE 
[JULY 22, 1915 

RECENT PROGRESS IN PYROMETRY.} 
ps the past five years, which is the period 
intended to be covered by the present paper, 
considerable advances have been made in the produc- 
tion of instruments for the measurement of high 
temperatures. Much valuable worl in this direction 
has been carried out at the National Physical Labora- 
tory in this country, and at the United States Bureau 
of Standards; and the manufacturers of pyrometers, 
chiefly in Britain and America, have introduced many 
new instruments for scientific and industrial purposes. 
Owing to the claims of atomic and molecular physics, 
the subject of pyrometry has not received that atten- 
tion from physicists in general to which it is entitled 
by its industrial importance. The progress made in 
various directions may conveniently be considered 
under separate headings. 
Standards of Temperature. 
Various investigations of fixed points have con- 
firmed, or at the most slightly modified, previously 
accepted figures. Up to the highest reading obtain- 
able on the gas scale (1550° C.), the standards now 
in use appear to be well established; and beyond 
this the melting-point of platinum, as deduced by 
several different methods, is now accepted as 
1755° C. This is a useful fixed point for the calibra- 
tion of high-reading pyrometers. 
The United States Bureau of Standards now issues 
materials of certified fixed points for the calibration 
of pyrometers. This procedure might with advantage 
be followed by our own National Physical Laboratory, 
so as to enable the indications of pyrometers to be 
checked from time to time by the user. If accom- 
panied by instructions for use, a correct result would 
be ensured, and the danger of error resulting from 
the employment -of materials of doubtful purity 
eliminated. 
Thermo-electric Pyrometers. 
One of the chief features of recent years has been 
the extension of the use of base-metal junctions in 
place of platinum and platinum alloys. Suitable base- 
metal couples develop a relatively high E.M.F., and 
enable a strong and cheap indicator to be used in 
place of the sensitive instrument required for couples 
of the platinum. series. Most makers now furnish 
iron-constantan pyrometers, which may be used up 
to goo® C. R. W. Paul employs two iron-nickel 
alloys of different composition which will register 
1o00° C.; whilst the Foster Instrument Company use 
two nickel-chromium alloys, capable of reading to 
1200° C. in continuous use, and to 1300° C. for occa- 
sional observations. For temperatures below 700° C. 
copper-constantan junctions are much used, as, for 
example, in  pyrometers for superheated steam. 
Various other base-metal couples are in use. 
The trustworthiness of base-metal junctions has 
been called into question by Kowalke (Transactions 
of the American Electrochemical Society, vols. xxiv. 
(1913) and xxvi. (1914)). Trials made on junctions 
of this type, as supplied by American makers, showed 
changes in calibration on heating for twenty-four 
hours ranging from 20° C. to 130° C., the higher 
figure being obtained at temperatures of 1000° C. 
The results showed the necessity of ‘‘ageing’’ the 
junctions by prolonged heating at the maximum tem- 
perature prior to calibration. Experience with British- 
made junctions shows that large errors of this nature 
do not occur, owing to a careful choice of materials 
and suitable treatment before calibrating. 
1 Abstract of a paper read before the Royal Society of Arts on May 12 b 
Mr. Chas. R. Darling. tt y M 
NO. 2386, VOL. 95] 



C. C. Bidwell (Physical Review, June, 1914) has 
shown that a junction of carbon and graphite may be 
used to read temperatures as high as 2000° C, 
Previous heating to this temperature is necessary 
before calibrating in order to expel volatile matter, 
This junction gives promise of a valuable extension 
of the range of thermo-electric pyrometers, and may 
find industrial applications. 
Materials for protecting junctions from the corro- 
sive action of furnace gases have been added to by 
the introduction of alundum (oxide of aluminium), 
which melts at 2050° C.; a material known as 
“silit,’ which has a carborundum basis; and 
‘ silfrax,’? a substance resembling carborundum. All 
these materials are brittle, and therefore will not 
stand rough usage. As a protection for junctions 
used to read the temperature of molten brass or 
bronze, a tube of molybdenum has proved successful, 
as this metal is a good conductor of heat and is not 
acted on by the molten alloy. 
Indicators for thermo-electric pyrometers have been 
improved in details by the various malsers, resulting 
in greater trustworthiness. .\ new departure in com- 
mercial indicators has been made by the Leeds and 
Northrup Company of Philadelphia, who have adopted 
the potentiometer principle, formerly used only for 
accurate laboratory work. The connections are shown 
in Fig. 1, where B is a 2-volt accumulator, R, an 


ue 
. Pye 

A 
Fic. r.—Potentiometer indicator. 
adjustable resistance, R, a fixed resistance, DE a 
uniform stretched wire, S a standard cell, G a sensi- 
tive galvanometer, and A a switch. In order to 
maintain a steady fall of potential along DE, R, is 
adjusted so that on switching the standard cell into 
the circuit of the galvanometer, no deflection is ob- 
served. As the voltage of B falls off with use, this 
adjustment must be made from time to time. The 
pyrometer is connected to one terminal of the galvano- 
meter and to a sliding contact F, which, in taking a 
reading, is moved along DE until no deflection is 
observed on G. From the known relation between 
temperature and E.M.F. for the junction used, DE 
may be graduated so as to read temperatures directly. 
The advantages of this arrangement are greater sensi- 
tiveness and independence of the resistance of the 
pyrometer and leads; the drawback, from an indus- 
trial standpoint, is that the readings are not auto- 
matic. Several methods for adapting an indicator 
for special ranges have been devised. In one form, 
due to the Cambridge Scientific Instrument Company, 
the indicator takes the place of the galvanometer G 
in Fig. 1, and by fixing the slider F at a given 
position—representing, say, 500°—the pointer of the 
indicator is prevented from moving until this tem- 
perature has been reached by the junction. The zero 
of the indicator is thus made to represent 500°, and 
