produced by varying Magnetic Induction. 475 



I also proceeded to construct a larger and more carefully 

 made iron ring, instead of a mere hank as bought. I thought of 

 making a f5gure-of-8 ring, winding one loop with wire, and 

 hanging the needle in the other loop ; but practical difficulties 

 postponed the adoption of this plan, a circular ring being so 

 much easier ; so I went on with the notion of winding the 

 copper wire only over a sector, and screening that by copper 

 sheet. To make this ring an enormous quantity of fine iron 

 wire was procured, and wound for a great part of a month 

 on a suitable wooden mould in a slow-motion lathe, the wire 

 being passed through a flame and through a stick of shellac on 

 its w^ay, so as to roughly insulate its turns, in case we wanted to 

 employ rapid reversals. 



The winding continued until the wooden mould began to 

 break down, when it was extracted, all matted together, and, 

 mounted on a stone pillar standing on rock in the laboratory. 



Some natural difiiculty is found in getting the charged 

 needle to hang as wished, with its mirror in the right aspect. 

 Some difficulty is also found in arranging that the stability of 

 the needle shall not be too much increased by its electric charge, 

 so that it oscillates violently about some point and is not pro- 

 perly slow and sensitive. 



Another and very tedious difficulty is to arrange so that the 

 position the needle is willing to take up before and after charge 

 is nearly the same ; for if the two be very different the spot 

 of light will not keep still, but will sail steadily along as the 

 charge slowly leaks away. To detect small effects the spot of 

 light must be very still. It was difficult to secure this without 

 at the same time applying too severe a constraint, either electric 

 or other, to the needle. It had to be done pretty much by 

 selecting carefully the shape to be given to the curved plates 

 of the needle, and to the enclosure opposed to them. 



All these difficulties were gradually more or less overcome 

 through the patience and skill of my laboratory assistant, 

 Mr. Benjamin Davies. 



There remained a few irregular disturbances, some of which 

 could be traced to convection-currents, others to the ordinary 

 movements about a building, and others again to the passage 

 of London and North-Western trains in their tunnel, some 

 150 yards aw^ay and 60 feet down, in the sandstone rock. 



The laboratory at Liverpool is very favourabl}^ situated as 

 regards shaking. It is a substantial stone building, with 

 walls 2 feet thick, on sandstone rock, which is partially isolated 

 from the railway by the remains of an old quarry, which has 

 been filled with rubble. Street traffic is very distant, and has> 

 never been appreciably felt to my knowledge. Students 



