HYDROGRAPHY. ,r 



liquid. Thus a mixture of distilled water, salt water and sulphuric acid makes its way from B, into 

 B 2 , and diluted potash from B 2 to 2? 3 , which beforehand contains pyrogallic acid & a solution of po- 

 tassium hydrate. As soon as the gas is transferred from one burette into the next one, the former 

 is ready to receive another quantity for absorption and measuring. At the same time a constant 

 renewal of the water in the burettes is taking place from the water-receiver L,; but on account of 

 the water being nearly saturated with air during the whole of its passage, the quantity of air that 

 is lost by absorption at improper places, will not be great in so far as the quantity of oxygen and 

 nitrogen is concerned. On the other hand, the absorption of the carbonic acid in the burette B t may- 

 be the cause of some inaccuracy, but I preferred this to any further complication of the apparatus, 

 as I did not consider the determination of the absolute quantity of carbonic acid to be of much 

 consequence. 



Being of opinion that a generation of hydrogen was necessary, when the carbonic acid was 

 to be boiled out completely, I carried out this object by placing a piece of zinc in the mercury. The 

 acid that was added produced then a generation of hydrogen. It was not, however, always conve- 

 nient to be obliged to take regard to this generation of hydrogen, and, moreover, it would make the 

 determination of the nitrogen unreliable. Owing to this, I made up my mind to avoid it entirely by 

 removing the zinc, as I had come to the conclusion that this little modification was to be preferred, 

 even if it should be at the sacrifice of the exactness of the determination of the carbonic acid. 



It is to be observed, however, that the quantity of heat imparted to the boiling-flask is very- 

 great, and that the cooling of C is very energetic, in consequence of which the ebullition becomes 

 specially violent towards the completion, even attended with shocks, but this is a matter of no conse- 

 quence at all. 



The burettes were divided with great exactness, but they were calibrated for all that after 

 the determination of the zero-point had taken place with great exactness; by this every single reading 

 can be subject to correction. Each single reading of volume has, by means of this correction, been 

 reduced to a pressure of o° and 76o n,m , and the dryness according to the table on page 30 — 32 in the 

 Landolt & Bornstein «tables». The application of this table makes it a condition that the height of 

 barometer should be measured with a mercury barometer, but it can easily be seen that the error 

 occasioned by the exclusive use of aneroid barometers, which are those I have used, is hardly 

 perceptible. 



That the mercury can be removed, which possibly might get into the burettes B l and B 2 , 

 these latter are at the nethermost part closed with a little tube and a jam-cock. This precautionary 

 step facilitates also the cleaning of the apparatus, but, according to what has been said above, a 

 purification of this kind is very seldom necessary, for the mercury is for a time collected in the glass 

 tubes in which the burettes terminate below, and it cannot run into the hoses that connect the 

 burettes with the water-receivers L. We have thus, when the water-surface in L is kept on a level 

 with the surface in the burette, with sufficient exactness the pressure of the atmosphere on the 

 quantity of air that is to be measured. 



The apparatus can be used with as great an exactness, and nearly with as great a facility 



5* 



