234 Mr. J. B. Hannay's Examination of Vacua. 



used for H. c and b are now opened, and the metal run out 

 of E, which is again heated, and new metal passed round 

 from D, which holds sufficient metal for six fillings of E. 

 The residual air, if there is any, is always imprisoned by a 

 new layer of frozen metal in H. The bulbs were then sealed 

 off; but even after all this treatment, a little air was seen to 

 come off when quite melted. The best evacuation obtained 

 in this way was tested as follows. One bulb was placed in a 

 freezing-mixture of calcium chloride and snow, and the other 

 interposed between a standard candle and radiometer kindly 

 lent me by Sir TV. Thomson; but no retarding effect further 

 than that due to the glass was observed, as on admitting air to 

 the bulbs no acceleration took place. 



It was found that, by increasing the proportion of lead in 

 the alloy by 20 per cent., the expansion on solidification was 

 greatly reduced, and enabled the freezing method of stoppage 

 to be used : this raised the melting-point to 108°. The last 

 apparatus was now modified, so as to dispense with stopcocks, 

 and a new attempt made. The bulbs were also reconstructed for 

 testing the conducting-power of the vacuum for electricity, one 

 bulb being made as shown in fig. VII. In this the platinum 

 wires were led into cups filled with fusible metal, so that no 

 leakage along the platinum wire was possible. Wires were 

 led into the cups from the outside, insulated with asbestos 

 and shellac, and the spark from a strong Holtz machine made 

 to pass. At the highest exhaustion obtained by the Sprengel 



P nm P> 23 ooo ooo °f an atmosphere measured by the McLeod 

 gauge, the spark passed as a flash illuminating the glass 

 surface inside ; but on exhausting as above with fusible metal, 

 no power I could obtain would cause a spark to pass through 

 the vacuum. A Leyden battery of 12 jars was tried, and a 

 coil giving a 12-inch spark, but with no effect. Mercury 

 vapour seems, therefore, to conduct the spark. An attempt 

 was made to detach the bulbs, but they were fractured. It 

 was found that it was by no means necessary to keep the 

 fusible metal in a vacuum, as it dissolves no gases ; and as the 

 great length of the apparatus rendered its construction trouble- 

 some and its fracture easy, a new apparatus was made as 

 shown in fig. VIII. 



This was constructed to obtain a simple torricellian vacuum, 

 a movable iron bucket below being provided, so that the 

 metal could be raised and lowered. The tube, from the bulbs 

 downwards, was 48 inches long ; and the lower end was turned 

 into a spiral, with the termination turned inside the spiral, so 

 that the end could nowhere touch the tube. I need not describe 



