1895.J Is there Argon in Vegetable or Animal Substances? 491 



potash solution, whence it was transferred to a small gasholder 

 containing water which had been boiled for some hours. The gas- 

 holder was connected with an apparatus similar in general arrange- 

 ment to that used by Professor Ramsay in July last, which enabled 

 the gas to be dried and purified from hydrocarbons, carbon monoxide 

 or hydrogen, the nitrogen being absorbed by being passed and 

 repassed over magnesium turnings kept at a temperature of about 

 600 C. A three-way stopcock enabled the magnesium tube and the 

 purifying and drying tubes adjoining it to be connected either with 

 the second gasholder or with a " Sprengel," so that any residual gas 

 could be pumped off, and the tubes exhausted before admitting the 

 gas. The gas pumped off was collected in a piece of hard glass 

 tubing ; it was then mixed with oxygen prepared by heating potas- 

 sium permanganate, and sparked down until no further diminution 

 of volume occurred, when excess of oxygen was known to be present. 

 The remaining gas was transferred by a gas pipette to a tube standing 

 over mercury, and the oxygen present absorbed by potassium 

 pyrogallate solution. Only one experiment was carried through to 

 the end in each case as the results seemed conclusive, although 

 several determinations of nitrogen were made. 



Peas were selected as a typical vegetable, and the following is a 

 summary of the results : 



342 c.c. of the volume collected in the first experiment (which was 

 done in two parts) was passed over the heated magnesium until so 

 much nitrogen was absorbed that both gasholders were full of water, 

 and the residue was then left for about 12 minutes in the tubes, the 

 magnesium being kept heated so that absorption of nitrogen might 

 continue. On connecting with the pump, a pressure of about 

 2 in. was registered, and the volume of gas pumped off had a 

 volume of nearly 8 c.c. After sparking down with excess of oxyge?i 

 6'8 c.c. of gas was left, and on absorbing by a strong solution of 

 potassium pvrogallate (stronger than Hempel recommends for com- 

 plete absorption) 0'12 c.c. of gas remained, which is less than 0*04 

 per cent, of the volume taken. To see if this bubble of gas consisted 

 of carbon monoxide, two experiments were performed with the oxygen 



