Velocities ofu Particles from Radioactive Substances. 559 



a'Hda: and I {l 2 — cc)Hdx 

 o Jo 



could have been obtained directly by a method given by 

 Sir J. J. Thomson*, in which the magnetic flux through 

 triangular coils of appropriate dimensions is measured on a 

 ballistic galvanometer or fluxmeter; but in the present case 

 it was judged to be more convenient and more accurate to 

 investigate the small variations of the field from point to point. 

 The field at the centre was measured carefully with a search- 

 coil and ballistic galvanometer, and the variation of the field 

 tested by moving a small search-coil rapidly from the centre 

 to different points. The values of the integrals are then most 

 conveniently obtained by plotting the graphs of the variation 

 of the field with distance from the centre, and from the area 

 enclosed by the curves deducing the small correction to be 

 subtracted from the value of the field at the centre to give 

 the mean effective field H. 



A few particulars are given of the construction of the 

 search and the standardizing coils which were specially 

 made for the present experiments. The standard search-coil 

 emploj^ed consisted of nine turns of silk-covered No. 42 

 copper wire, wound on an ebonite disk of 3*59 cm. diameter; 

 the small coil used to investigate the variation of the field 

 had 46 turns of No. 47 copper wire on a disk of 1*59 cm. 

 diameter. The primary and secondary solenoids required 

 for the standardization of the fields were each made of a 

 single layer of wire, so that all the data required could be 

 obtained with ease and accuracy. The primary consisted of 

 600 turns of No. 16 copper wire wound on a brass tube 

 1 metre long and 5'8 cm. diameter ; the coil was wound very 

 carefully on the lathe, and the distribution of the windings 

 afterwards investigated over the whole length with a reading 

 microscope. To compensate for the smallness of the magnetic 

 field obtained with reasonably small currents in such a solenoid, 

 it was necessary to have the secondary coil of a large area: 

 the secondary employed had a total area of J 0,200 sq. cm., 

 consisting af 655 turns of No. 42 copper wire, of mean 

 diameter 4'45 cm., and had a resistance of about 200 ohms. 

 This was wound on a piece of wide ghiss tubing, selected for 

 its uniformity of bore. The variation in the sectional area of 

 the tube was examined by measuring at intervals of 1 cm. 

 along its axis, and after the coil had been wound, the corre- 

 sponding density of the windings at each section was obtained 

 by counting the turns under the reading microscope. The 

 secondary solenoid was held centrally within the primary, 

 * Sir J. J. Thomson, Phil. Mag. xviii. p. 844 (1909), 



