Mossi'?. G.J. StoiU'V (i/id li. J. Moss oaa Crooke.<'s Force. ()9 



Spreiigel pump. The pump was set in act ion, and occasionally th(^ 

 fiame of an ordinary gas-burner was held at a distance ol:' about 

 10 centims- from the blackened pith while the microscope-glass 

 ^vas closely watched. When the gange of the pump showed a 

 tension of 7 niillims., as compared v^ith the mercurial column of 

 a barometer standing in the same vessel of mercury, the glass disk 

 was distinctly repelled from the pith and towards the source of 

 light. As the exhaustion was continued the repulsion between 

 the pith and the glass increased. The apparatus was sealed off 

 from the pump when the mercury falling in the exhaust-tube had 

 for some days produced a metallic sound. Feeble illumination 

 now caused the glass disk to be forcibly driven away from the 

 pith*. 



We now endeavoured to determine quantitatively the influence 

 of variations in the tension of the residual gas, and also the 

 influence of variations in distance between the reacting surfaces. 

 For this purpose we constructed the apparatus represented in 

 fig. 2. 



On a wooden stand supported by three levelling-screws rests a 

 glass tube 20 centims. in length and 3*8 centims. in diameter, 

 having a tubular opening at one side, into which is cemented hori- 

 zontally a smaller tube 1*5 centim. in diameter. In the larger 

 tube there is a circular disk of elder-pith 2*3 centims. in diameter, 

 having one side blackened with lampblack ; it is supported in a 

 vertical position on a movable stand of iron wire. By means of 

 a magnet the pith disk can be moved up and down the tube, and 

 thus placed at any required distance (within 12 centims.) from 

 a delicately suspended circular disk of thin microscope-glass, 

 3 centims. in diameter and 0*3 millim. in thickness. The glass 

 disk is attached to the end of a glass arm, which is suspended 

 in the smaller tube by means of a silk cocoon fibre contained in a 

 vertical limb 38 centims. in length and 9 millims. in diameter. 

 In order that the torsion of the silk fibre may be conveniently 

 regulated, there is a small fl-shaped piece of iron wire attached 

 to it a few centimetres below the end from which it hangs, A 

 horseshoe magnet is suspended outside the tube with the piece 

 of wire between its poles. By turning the magnet round, torsion 

 maybe imparted to the silk fibre. The balance of the glass arm is 

 adjusted by means of a sm.all iron ring which it carries; the position 

 of the ring can be altered at will by an external magnet. There 

 ia a small silvered mirror attached to the arm at the point of sus- 

 pension; this reflects the image of a narrow illuminated slit onto a 

 scale divided into degrees 2*5 millims. each. An alteration in the 

 position of the index amounting to 0*5 millim. is readily observed ; 

 this corresponds with a change in the position of the outer edge of 

 the glass disk amounting to 0*033 millim. One end of the large 

 tube is ground perfectly flat, and closed by cementing to it a plate 



^ The apparatus was sealed cif on the 14th of April, 1876. The experiments 

 described above were made in March. 



