326 MESSES. C. CUTHBERTSON AND E. B. K. PRIDE AUX 



combine with an element whose fluoride was a solid (lead was chosen), measuring 

 the contraction of volume so produced by means of mercury manometers which were, 

 however, kept as far as possible from the fluorine by a column of air. 



In the figure, Lj and L 2 are graduated glass tubes of the same bore, having, at 

 their upper extremities, narrow tubes filled with dry lead filings, K a , K 2 , and 

 terminating in platinum tubes, Pj, P 2 , which fitted the leads of the refractometer 

 tube. These closed burettes were connected with open movable burettes, Mj, M 2 , 

 also of equal bore, which were joined by a wire passing over a pulley. One of the 

 burettes (L 2 ) was in connection with a graduated reservoir of mercury, R, provided 

 with a tap, and all were filled, to the proper point, with dry mercury. 



When the air in the refractometer tube had been displaced by the gaseous products 

 of electrolysis as completely as possible, and the number of bands which had crossed 

 the field had been noted, the entry tube was disconnected, and the two tubes 

 P!, P 2 rapidly connected with the system of burettes, the junctions being made air- 

 tight by immersing them in mercury. The burette M 2 was then raised slowly, while 

 M 1; being connected by a pulley, fell by an equal amount. The mercury in LI, L 2 

 followed the motion of that in their respective companions, and the effect was to push 

 the contents of the refractometer tube into the glass tube at Kj, which was filled 

 with lead filings, without appreciably altering the pressure, so as to avoid errors due 

 to possible leaks. As the fluorine combined with the lead, there took place a 

 diminution of the volume of gases in the closed space K 1; H, K 2 , which was indicated 

 by a difference of level between the mercury in the closed and open burettes. As 

 fast as this was observed, pressure was equalised by opening the tap Q and letting in 

 mercury. By continuing and reversing this process the gases in the refractometer 

 tube were pushed back and forward for about an hour. When no further change of 

 levels could be detected it was assumed that all the fluorine present had been 

 absorbed, and that the residual gases consisted of oxygen and nitrogen. The volume 

 of fluorine was measured by observing the change of levels of the mercury in the 

 two closed burettes, and the measurement checked by reading the change of level in 

 the reservoir R. 



The amount of oxygen present was found by taking a sample of the total residual 

 gases in the closed system and burning it with phosphorus. The residue was also 

 tested for SiF 4 , and, finally, was shown by its index and spectrum to be nitrogen. 



From these data the index of fluorine could be found. But as the calculation is 

 rather long, a specimen is given below, the figures being those actually observed in 

 the third experiment given in Table III. 



