294 CARNEGIE INSTITUTION OF WASHINGTON. 



Gaugain galvanometer (see below); the construction of coils, termi- 

 nals, and core will be according to the design perfected by the National 

 Physical Laboratory. 



Two coils, wound on well-seasoned wood rings, were mounted at one 

 of the stations in the standardizing magnetic observatory, with suit- 

 able direct-current circuits and sliding resistances, for the purpose of 

 producing different horizontal-intensity fields for the calibration and 

 testing of variometers. These coils are about 1.8 meters in diameter 

 and are spaced a distance apart equal to their radius on a non-magnetic 

 frame arranged to permit vertical adjustment of the center of system 

 to any height between 8 cm. and 25 cm. above the top of the pier. 

 This adjustment makes possible the testing of any instrument the 

 essential magnetic unit of which is within 8 cm. to 25 cm. of the sup- 

 porting base. 



WORK OF DIVISION OF EXPERIMENTAL WORK. 

 MAGNETISM. 



Improvement in the design of a sine galvanometer. — This instrument, 

 whose construction has been begun, has been designed to give, with 

 much less expenditure of time than that necessary with magnetometers, 

 a precision greater than that obtainable with them or with other sine 

 galvanometers or other electrical devices hitherto constructed for the 

 purpose. The essential parts of the instrument are the coils, the mag- 

 netometer, and the circle, together with their adjuncts. 



The coils, arranged as in the Helmholtz-Gaugain galvanometer, are 

 to be wound on the best white statuary marble. Each coil will consist 

 of an exact integral number of turns of thin bare copper wire wound 

 under tension in a single layer in a lathe-cut spiral groove. Each coil 

 will be wound in halves for insulation testing. The design is such that 

 it should be possible to determine the constant of the coils to 1 part in 

 37,000 at least. The error due to a radial displacement of the magnet 

 by even 5 mm. would be only 1 part in 20,000, and that due to an axial 

 displacement of 1 mm. would be about 1 part in 40,000. 



The magnetometer will be a simple one, with small, fiat steel mag- 

 net-mirror, marble house, and quartz suspension. Two circles are 

 provided, both of large diameter and one of them finely divided. The 

 error from circle reading can probably be made not greater than 1 part 

 in 40,000 or 50,000. The electrical measurements can be made with 

 adequate precision, probably to 1 part in 30,000. 



The instrument is to be used both in the usual way and also in the 

 way recently proposed by Schuster (Terr. Mag., 19, 1914, p. 19). 



Investigation of certain aspects of crystal magnetization. — Three pieces 

 of apparatus have been designed and for the most part completed for 

 this work, but the difficulty of getting suitable material has prevented 

 useful results from being obtained. The most important part of the 



I 



