346 



The gas-burette must always contain a litle moisture, 

 but care should be taken to avoid water in the narrow 

 graduated part. 



Before a series of С O2 -determinations the apparatus must 

 always be tested by passing a sample of air several times to 

 and fro between the burette and the absorption-pipette and 

 seing that the volume does not alter. If it does, the apparatus 

 is either not air-tight, or else the soda-lye absorbs or gives off 

 gases. This latter is not unfrequently the case when the 

 temperature or the barometric pressure, or both, have varied 

 considerably since the last analyses. If the fluid is not com- 

 pletely saturated with atmospheric air, at the pressure and 

 temperature obtaining, it will give off or absorb (according to 

 the circumstances) minute quantities of air during each analysis 

 and thereby vitiate the results. Freshly prepared solutions of 

 sodium-hydrate are always greatly supersaturated with air. The 

 best remedy is to draw off the fluid into a large bottle 

 and shake it vigorously for some time before recharging the 

 apparatus. 



If, as in my apparatus, the two bulbs of the absorption- 

 pipette are connected by rubber-tubing it is very easy to empty 

 them, but the rubber introduces a new source of trouble 

 viz., that the soda-lye may in the course of time dissolve a 

 little sulphur and thereby acquire the power of absorbing 

 minute quantities of oxygen. A slight yellowish tint indicates 

 this state of affairs and such solutions must be immediately 

 renewed. 



If the narrow part of the burette is graduated, as it ought 

 to be, in Vioooo parts the difference between the two readings 

 gives the tension directly, and it is unnecessary to apply correc- 

 tions even if the original volume is not exactly 10000. If great 

 accuracy is desired, and especially if the determinations are 

 made at greatly varying temperatures, it becomes necessary to 

 make allowance for the varying tensions of water-vapour and 



