486 PROFESSOR KNOTT ON SOME RELATIONS BETWEEN 



Gore, or myself, the wire or rod that was being studied was hung vertically inside a 

 magnetising helix. In such cases, the effect of the earth's vertical field had to be allowed 

 for. Certain discrepancies in Gore's observations may, I think, be explained in terms of 

 the effect of this permanent downward field, acting with or against the field inside the 

 helix, according to the direction of the magnetising current. 



The most satisfactory way to get rid of this permanent effect is to have the wire 

 perpendicular to the earth's field. At the same time it must be left free to twist. These 

 conditions were completely fulfilled as follows. Two identical magnetising coils were set 

 horizontally in line, and perpendicular to the magnetic meridian. The wire to be 

 examined passed through both coils, and carried a small mirror fixed to its middle point. 

 The contiguous ends of the mognetising coils were separated a few centimetres to allow 

 of the adjustment by hand of the mirror on the wire. A vertical scale was reflected from 

 the mirror and read by means of a cathetometer telescope suitably adjusted. The current 

 was led through both coils put on in series in such a way that the coils formed two 

 electromagnets oppositely directed as regards polarity. Thus the iron or nickel wire 

 that threaded the helices was magnetised in two halves, so that its middle was one kind 

 of polarity, and its extremities the other kind. When a current was passed along such a 

 wire, the two halves were subjected to opposite twists ; and, since the extremities were 

 clamped, the middle point of the wire with its attached mirror rotated through the angle 

 corresponding to the twist. It seems reasonable to suppose that the angle of rotation of 

 the mirror in this arrangement will be very nearly the same as would be produced if the 

 half length of wire were hung vertically, and subjected to the same combination of 

 longitudinal and circular magnetic forces. It is possible, however, that the different 

 conditions of constraint in the two methods may give rise to differences in the twist 

 produced by a given combination of magnetisms. There are indications of such 

 differences, but they are not great. 



The tension in the wire was applied by means of a screw at one end, and was measured 

 on a spring dynamometer at the other. After a given tension was applied, the wire was 

 lifted slightly on two brass edges just outside the coils at the further ends of the cores. 

 A free length of wire, 77 cm. long, was thus kept stretched from end to end of the double 

 coil. Of this length 7 '5 cm. lay in the region between the coils — that interval being 

 necessary so as to admit of easy adjustment of the small central mirror. 



In both the coil and line circuits galvanometers were included, and were placed 

 sufficiently distant from the coils and from one another so as to prevent any electro- 

 magnet disturbances. The mercury commutators, by which the currents were made, 

 broken, and reversed, were placed within easy reach of the observer, who sat with his 

 eye at the cathetometer telescope. 



3. The most interesting result brought out by these experiments is that the twist 

 produced by a given combination of circular and longitudinal magnetisms depends largely 

 upon the order in which these are applied. This peculiarity was mentioned in the last 

 paragraph of Part I. Gore has also recorded certain observations which indicate the 



