12 SCIENCE 
on the absolute measurement of current, 
using an improved form of Rayleigh cur- 
rent balance, has been thorough and ex- 
haustive. To obtain a final result reliable 
to two or three parts in a hundred thou- 
sand requires that all possible sources of 
error as great as a few parts in a million 
must be examined and taken into account. 
It would be impossible in this form of cur- 
rent balance to measure directly the dimen- 
sions of the coils with the required preci- 
sion, and hence the constant of the instru- 
ment must be determined by electrical 
means without such direct measurements ; 
and to detect and eliminate unsuspected 
errors, several sets of fixed coils and sev- 
eral different moving coils were prepared 
and used interchangeably, giving the equi- 
valent of several balances. This, in fact, 
did reveal unsuspected sources of error, 
and although it greatly prolonged the in- 
vestigation, it gave results far more trust- 
worthy in the end. A similar story could 
be told of the work at the bureau on mer- 
eury ohms, and on the determination of the 
ratio of the electrical units. 
These are some of the researches in con- 
nection with electrical standards, which 
together constitute the third group of in- 
vestigations in Division I. of the bureau. 
Some of them are described in papers con- 
tained in the Bulletin, and some are in 
press, and some are not yet completed. 
(d) Improvements in Instruments and 
the Development of New Instruments.—In 
addition to new instruments developed and 
improvements made in existing instru- 
ments employed in the absolute measure- 
ments mentioned above, many improve- 
ments have been made in electrical instru- 
ments used for other kinds of electrical 
measurements. 
Some of these improvements have been 
adopted by American and foreign instru- 
ment makers, and some are embodied only 
[N.S. Vou. XXXV. No. 885 
in instruments in our own laboratories. 
Among these may be mentioned the follow- 
ing: 
(1) The Bureau of Standards type of 
sealed resistance standard, which is used 
not only for resistances of highest precision 
in standardizing institutions, but also for 
precision standards in ordinary use. 
(2) The Bureau of Standards chrono- 
eraph, for measuring and recording with 
extreme precision the speed of a machine, 
as, for example, the speed of a dynamo for 
the purpose of obtaining the frequency of 
the current generated with highest pre- 
cision. It is used on the new apparatus 
for the measurement of absolute resistance 
at the bureau, and in connection with the 
absolute measurement of capacity and in- 
ductance, and has been adopted by the Na- 
tional Physical Laboratory on the magnifi- 
cent machine recently erected at Tedding- 
ton. 
(3) The direct reading potentiometers 
designed for rapid and accurate work in 
the measurement of current and voltage, 
which have been of great service in the 
work of the bureau, and would be more 
largely used outside if instrument makers 
had been quicker to appreciate their merits. 
(4) The series of electrodynamometers 
for the measurement of alternating cur- 
rent and power, which can be calibrated 
by direct current and used on alternating. 
Their range is up to 1,000 amperes, and 
one recently constructed, but not yet fully 
installed, will have a range of 5,000 am- 
peres. They are essentially instruments 
for the testing laboratory, and without 
them the bureau would have been unable 
to make many of the tests which have been 
successfully carried out. 
(5) Important improvements have been 
made in instruments for the accurate com- 
parison of resistance standards and the 
testing of resistance boxes, potentiometers, 
