322 CARNEGIE INSTITUTION OF WASHINGTON. 



and the choice of a coil of such a form that the field at its center and the 

 variation of the field over small distances are capable of being accurately 

 calculated. When the coil is properly chosen, it would appear that the field 

 due to it can be determined to a degree of accuracy surpassing that usually 

 required for the determination of H, or at any rate surpassing that obtained 

 in the same time by the magnetic method of measuring the field. The constant 

 of the coil should, in a properly constructed one, be capable of being known 

 at least within two or three parts in 10^, and, with the modern refinements, a 

 corresponding accuracy in the current measurements seems readily possible. 

 The accuracy of the current measurements, so far as the observer is concerned, 

 depends on the accurate knowledge of a standard resistance, and of the e. m. f. 

 of a standard cell. The proposed methods, by which the field due to the coil 

 is compared with that of the Earth, fall into two classes: (1) methods involving 

 the measurement of a deflection, the coil being used in fact as a galvanometer; 

 (2) methods in which the field due to the coil is compared with the Earth's 

 field by balancing it directly against that field, the comparison being made by 

 some appropriate method. 



In the first of these classes belongs Watson's method,^ the principle em- 

 ployed being that of the sine galvanometer. In 1912 the problem of measure- 

 ment of the Earth's magnetic field by electrical methods was assigned to Dr. 

 N. E. Dorsey, while he was connected with the Department as Research 

 Associate. As the result of his study and his extensive experience in electrical 

 measurements in the United States Bureau of Standards, he likewise finally 

 chose, as did Watson, the absolute sine galvanometer for measuring H, the 

 horizontal intensity. Dorsey preferred, however, the simple coil instead of 

 the Helmholtz arrangement and, of course, proposed measuring the current 

 according to the most approved present-day methods. He made a thorough 

 investigation of the theory of the instrument, examined into the various 

 possible sources of error, and reached the conclusion that an absolute accuracy 

 of 1 in 10,000 could be attained. Taking all things into consideration, a 

 diameter of coil of about 25 cm. was chosen by Dorsey, and the number of 

 turns and size of wire and other practical details of construction of the various 

 parts of the entire apparatus were carefully considered by him and the neces- 

 sary specifications and drawings were prepared. His investigation and designs 

 were completed by March 11, 1913, but the proposed apparatus, pending 

 completion of other studies and designs, has not yet been constructed. 



To the second class belongs the method recently carried out by W. A. 

 Jenkins, in Professor Hicks's laboratory at the University of Sheffield,- as also 

 the method proposed by Schuster.^ The methods thus far cited aim chiefly to 

 determine the horizontal intensity, H, though they can readily be adapted, in 

 general, also to measuring the magnetic declination, D. 



Respecting the absolute accuracy attainable with electrical methods, it 

 must be borne in mind that freedom from magnetic impurities in the material 

 of the parts concerned is as essential as in the present magnetometers. In 

 order to make the electrical method universally apphcable, it may be found 

 preferable to measure the total intensity rather than the horizontal or vertical 

 component. If the same instrument were capable at the same time of meas- 

 uring H, or the vertical component, Z, when suitable to do so, the inclination 

 could likewise be obtained. Accordingly, designs for a universal instrument 



'Watson, W.: A determination of the value of the Earth's magnetic field in international 

 units. Phil. Trans. R. S., Ser. A, vol. 198, pp. 431-462 (London, 1902). 

 "Phil. Mag., vol. 26, pp. 752-774 (Oct. 1913). 

 »Terr. Mag., vol. 19. pp. 19-22 (1914). 



