18 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No. 2 
present time,” while it vies with heat of chemical reaction as a conven- 
ient secondary standard. 
The present paper deals with some features of electric standardiza- 
tion which need attention in working to 30 per million. 
It seems proper to take as the standard form of electric heating the 
well known method of measuring the time, the current through the 
heating coil, and the voltage drop across it; hence, for instance, if the 
resistance of the coil is used directly, it is to be measured between 
points which would be suitable for the attachment of the potential 
leads in the standard method. It is often desirable to make the time 
of heating as short as 3 minutes. 
The special topics to be treated here are: I, the conduct of the de- 
terminations, and, II, the error from heat produced in the leads. In 
conducting the determinations it is necessary to deal with: A, the 
variety of observations, and B, the variations in the heating currrent.* 
I, A. Management of the Observations 
A second observer, with complete apparatus, can take care of the 
electric energy readings, including the time, so that the calorimeter 
can be observed as usual. A single practised observer, with good 
apparatus, can also do all the work alone if the calorimeter temperature 
is read electrically; most simply if a potentiometer can be used for all 
the measurements. A superior method, however, for fluid-filled calori- 
meters and heatings not over 5 minutes long, is to calculate the average 
calorimeter temperature during the heating, by means of observations 
made immediately before and after. It evidently makes the second 
observer quite superfluous, and enables a single one to give more com- 
plete attention to the energy readings. It therefore seems to deserve 
examination. 
2 To define the calory electrically, (as, say, 4.183 joules), is really only to recognize” 
a situation already existing. Such recognition, as soon as it becomes general, will put 
an end to the use of the needless multiplicity of ‘‘calories’’ (15-degree calory ; 20-degree 
calory, etc.) which now inconveniences us, since the only excuse for more than one 
standard is uncertainty as to the exact ratio between values determined by means of 
water at different temperatures. Few, if any existing results obtained in water-de- 
rived calories are accurate enough to suffer appreciable loss of accuracy through restate- 
ment in electric calories, while electrically derived results might often lose by the 
opposite transformation. 
All this is aside from the question whether heat is to be stated in terms of calories 
of approximately 4.183 joules, or in joules directly. 
® More difficult is knowing the values of the auxiliary resistance coils concerned in 
the energy measurement. This, however, has been adequately treated elsewhere. 
