THE SO-CALLED INTERNATIONAL ELECTRICAL UNITS."

By FRANK A. WOLFF.

As one of the most important questions likely to be considered by the St. Louis International Electrical Congress will be that of redefining the fundamental electrical units, it may therefore not be out of place to briefly review the efforts which have thus far been made to bring about international uniformity.

The need of a definite and universal system of electrical units was early recognized, and became a necessity as soon as industrial applications of electricity were made. At first the principal measurements were those of resistance (line resistance, insulation resistance, measurements for the location of faults, etc.). These were expressed in terms of some entirely arbitrary standard, such as the resistance of a given length of an iron or copper wire of given cross section. This naturally led to a great multiplicity of units, none of which ever gained general acceptance.

In 1848 Jacobi pointed out that it would be more satisfactory to adopt as a universal standard the resistance of a certain piece of wire, copies having the same resistance being easily constructed. Jacobi carried this suggestion into practice by sending copies of his standard, since known as "Jacobi's Étalon," to the leading physicists of that period.

In 1860 Werner von Siemens proposed as a standard of resistance the resistance, at 0° C., of a column of mercury of a uniform crosssection of 1 square millimeter and 1 meter in length.

In 1861 a committee composed of the most eminent English physicists was appointed by the British Association to consider the question of standards of electrical resistance. The leading foreign physicists were invited to offer suggestions, and various special investigations of the problems with which the committee was confronted were undertaken by its members.

a A paper presented at the International Electrical Congress, St. Louis, 1904.

It was decided that the unit of resistance should be defined in terms of the Gauss-Weber absolute system of electromagnetic units, which had already received such well merited recognition; but since this unit was inconveniently small it was decided to define the practical unit as an integral decimal multiple of the same.

The value of the unit depends upon the units of length, mass, and time adopted as the basis of the system. Those chosen by Gauss and Weber were the millimeter, milligram, and second. In England efforts were being made to establish an absolute system for the definition of all physical units, for which the fundamental units of Weber were of inconvenient magnitude, and for which the centimeter, gram, and second were finally adopted (the c. g. s. system).

The practical unit of resistance in this system was defined as 109 c. g. s. electromagnetic units, and while this definition fixes the unit theoretically, it can only be applied in practice by the measurement of some particular resistance in absolute measure. This requires the construction of especially designed apparatus, with which measurements lying within a very limited range may be made; the determination of its instrumental constants most frequently involving tedious mathematical approximations, and the elimination of errors of observation. With all possible precautions the errors of such methods exceed, even to-day, a hundred fold the relative errors in resistance comparisons.

Investigations were therefore made to determine whether the absolute unit of resistance could be accurately defined in terms of the resistance of a definite portion of a definite substance. The electrical properties of alloys and pure metals in the solid and liquid states. were studied with this end in view. On account of the excessive influence, on the resistance, of even small quantities of impurities in metals of the highest obtainable purity, and of small variations in the compositions of alloys, the choice was greatly limited. It was found, in addition, that solid metals had to be rejected on account of the marked influenced of physical changes produced by annealing, hardening, drawing, bending, etc.

Mercury, already recommended by Siemens, was therefore the only material to be further considered, but was also rejected for two reasons, viz, the large differences found to exist between coils supposedly adjusted to different German mercurial standards, and differences between a number of mercurial standards constructed by members of the committee.

The committee therefore recommended the alternative method of

constructing material standards adjusted with reference to the absolute unit. In this connection a special form of resistance standard known as the B. A. type was designed, and after an investigation of the constancy of a number of new alloys in addition to many already in use, one containing two parts by weight of silver to one part by weight of platinum was finally selected as best meeting all requirements.

In 1863 and 1864 the values of certain coils were determined in absolute units by one of the methods proposed by Weber, and from these measurements the B. A. unit was derived. A number of copies were issued, gratis, by the association, and in addition arrangements were made for supplying others at a moderate price. The B. A. unit soon gained general acceptance in the English-speaking countries, while the Siemens unit still retained its supremacy on the Continent.

No action was at that time taken by the British association committee to define the units of current and electromotive force further than in terms of the c. g. s. system. The currents to be measured were all relatively small, and were usually measured by means of a tangent galvanometer with a sufficient accuracy. Electromotive forces were seldom measured, and then usually in terms of the Daniell cell. In 1872 Latimer Clark brought to the attention of the committee the superiority of the cell which now bears his name, recommending it as a suitable standard of electromotive force, but no definite action was taken by the committee.

In 1878 it was shown by Prof. H. A. Rowland that the B. A. unit was in error by more than 1 per cent, and soon after the existence of a discrepancy of this magnitude was verified by a number of other investigators.

In 1881 a call was issued by the French Government for an International Electrical Congress, to be held in connection with the first International Electrical Exposition at Paris, for the purpose of adopting definitions of the electrical units which might serve as a basis for legislative enactments. In the meantime a number of mercurial standards had been constructed and had been found to be in satisfactory agreement; moreover, the results of most of the absolute determinations had been referred either directly or indirectly to the Siemens unit.

The Paris Congress, therefore, recommended that the practical electrical units be defined in terms of the units of the c. g. s. system of electromagnetic units, and that the unit of resistance be represented by a column of mercury 1 square millimeter in cross section, at the temperature of 0° C., of a length to be determined by an international

commission appointed for this purpose, as appears in the following resolutions:

RESOLUTIONS OF THE INTERNATIONAL CONGRESS OF ELECTRICIANS, PARIS, 1881.

(1) That the c. g. s. system of electromagnetic units be adopted as the fundamental units.

(2) That the practical units, the ohm and the volt, preserve their previous definitions, 109 and 10 c. g. s. units, respectively.

(3) That the unit of resistance, the ohm, be represented by a column of mercury 1 square millimeter in cross section at the temperature of 0° C.

(4) That an international commission be charged with the determination, by new experiments, of the length of the mercury column 1 square millimeter in cross section, at a temperature of 0° C., representing the ohm.

(5) That the current produced by a volt in the ohm be called an ampere.

(6) That the quantity of electricity produced by a current of 1 ampere in one second be called a coulomb.

(7) That the unit of capacity be called a farad, which is defined by the condition that a coulomb in a farad raises the potential 1 volt.

The Congress also recommended the employment of the carcel as the standard for photometric comparisons.

The international commission, appointed in accordance with paragraph 4 of the resolutions of the Paris Congress of 1881, met at Paris in 1882, but definite action was deferred until two years later, when the following definitions were unanimously recommended:

The legal ohm is the resistance of a column of mercury 1 square millimeter in cross section and 106 centimeters in length, at the temperature of melting ice.

The ampere is equal to one-tenth of a c. g. s. unit of the electro-magnetic system. The volt is the electro-motive force which will maintain a current of 1 ampere in a conductor of which the resistance is a legal ohm.

The value adopted for the length of the mercurial column was taken as 106 centimeters, notwithstanding that most of the best results were very close to 106.3; as it was thought advisable to adopt a value known to be true to the nearest centimeter for a period of ten years. On account of this uncertainty, no steps were actually undertaken by the various governments represented.

The conference also adopted as the unit of light of any color the quantity of such light emitted in a perpendicular direction by 1 square centimeter of molten platinum at the temperature of solidification; and as the practical unit of white light the total quantity of light emitted perpendicularly by the same source.

a For the sake of completeness, the recommendations of the various international electrical congresses on photometric standards are included in the summary.