of the Approach and Recession of Bodies. 77 



piece of platinum wire ending in a loop, which has been 

 cemented with gold, so as to leave no rough edge where the 

 silk fibre might become fast. The fibre is chosen of such a 

 length that when all is in position the balance will hang in 

 the centre of the bulb. The tube is now inverted, point 

 downwards, and the platinum wire first introduced at the 

 wider end, then the balance, supported by a wire with a 

 hook at the end, is carefully allowed to slide in until the 

 free end of the platinum wire emerges from the pointed eud 

 of the tube. The projecting wire is then caught hold of, 

 and the wire with the hook removed. The tube is now 

 inverted into position, and the balance adjusted so as to 

 swing in the centre of the bulb by raising or lowering the 

 wire at the top, before sealing it into the glass. The con- 

 traction towards the lower end, in anticipation of the 

 sealing off, is still further narrowed when it is ready for 

 attaching to the pump. After exhaustion it is sealed off. 



" My experience of the phenomena, as exhibited by the 

 arrangement described, may be stated very shortly. Although 

 indications of recession on the application of heat were 

 certainly obtained when the gauge was within eight milli- 

 metres (J inch) of the barometer, still these movements 

 were exceedingly feeble and uncertain. I found that the 

 exhaustion might reach to within -|=- of an inch of the 

 barometer, or even so near as to be indistinguishable from it 

 by the cathetometer, and yet when heat was applied no 

 certain indications of recession could be detected, nothing 

 further than an occasional slight twitching taking place, and 

 this even when a part of the glass was heated until painful to 

 the touch. I could only obtain indications with certainty 

 when the pump was made to work vigorously for some time 

 after the void was to all appearance as good as the barometer 

 vacuum. But so long as the pump was working, with 

 such a perfect vacuum, the behaviour of the balls under 

 the influence of heat was very erratic. The balance would 

 suddenly start off, making several revolutions each way, 

 slowing as the balls passed the flame, then swing to and from 

 the flame regularly for some time, soon to repeat these 

 spasmodic movements. 



" It was found on some occasions, when the Sprengel was 

 working under these conditions, that the approach of a stick 

 to within an inch or two of the glass caused one end of the 

 balance to swing towards and from the stick through 40°. 

 A test tube of hot water then caused the nearest ball to 



