about two-thirds of the shorter limb^ into which was then 

 introduced the amalgam after tiie latter had been wiped 

 with filtering paper. Then into the end of the limb con- 

 taining the amalgam, a caoutchouc stopper, perforated with 

 a small opening, was immediately thrust so far that its 

 upper surface came a little below the rim of the tube. The 

 decomposition of the amalgam was then allowed to proceed 

 for a few minutes, during which period any moisture ad- 

 hering to the amalgam or present in the tube would become 

 completely saturated with ammonia, and then the two gases 

 would begin to escape through the perforation in the stopper 

 in the proportions in which they are really evolved. Mercury 

 was now poured into the open end of the longer limb until 

 the amalgam just made its appearance at the top of the hole 

 in the stopper, which was then closed by pushing in a piece 

 of glass rod. The evolved gases being now retained in the 

 tube pressed up the mercury in the longer limb, and it was 

 from time to time drawn off by the outlet tube to prevent 

 undue pressure on the stopper. When the decomposition 

 was complete, which usually occurred in about IJ hours 

 (but in one case more than 2^ hours were required) the 

 mercury was brought to the same level in both limbs and 

 the space occupied by the gases was marked on the tube. A 

 little mercury was then let out so as to make the pressure 

 on the gas somewhat less than that of the atmosphere, and 

 the space above the stopper was filled with hydrochloric 

 acid diluted with a little water. The glass rod was then care- 

 fully withdrawn for an instant so that a few drops of the 

 acid might enter the tube. The ammonia gas present was 

 of course immediately absorbed, and the mercury having 

 been again brought to the same level in both limbs, the 

 space occupied by the residual hydrogen was marked on the 

 tube. The volumes occupied by the gases were determined 

 by finding the quantity of water required to fill them from 

 a burette. 



The following are the results of four experiments : — 



