194 Speedy Volumetric Determination of Carbonic Acid, 



after the absorption of carbonic acid, have risen to their original read- 

 ings, or very nearly so, the correction to be applied is all but limited 

 to the change of temperature in the 1600 c.c. of the absorption-tube r 

 and is hardly likely to exceed 4 or 5 c.c. 



The petroleum gauge is most interesting to watch ; when the ab- 

 sorption is complete, after the air has circulated three times through 

 the absorption-tube, the reading of the graduated bell-jar (absolutely 

 under atmospheric pressure) is extremely steady. It remains so for a 

 minute or two, during which time the thermometers rise in both bell- 

 jars ; this is owing to the circumstance that the glycerine in the tanks 

 is nearly always slightly colder than the water in the jacket of the 

 bell-jars, and the air on its way through the stand-pipe immersed in 

 the glycerine becomes a trine colder. As the temperature in the bell- 

 jars approaches its original reading, the pressure from the expansion 

 of the air they contain begins to assert itself, and in order to keep the 

 air under atmospheric pressure the graduated bell-jar must be gently 

 raised. When the thermometers have reached their original height in 

 the jacketed bell-jars of the instrument, the atmospheric pressure 

 again remains exceedingly steady, and there is no difficulty in reading 

 the result. It sometimes happens that after a time, especially in hot 

 weather, the temperature slowly rises by two or three tenths of a 

 degree or more beyond its original reading. In such a case the work- 

 ing of the instrument is so perfect that by introducing the correc- 

 tion required for this increased temperature in the proportion of 

 1 : 492 of the volume of air analysed per 1° P., the result obtained is 

 the same as it was on the occasion of the first reading. 



It is important to consider how far the instrument is reliable. At 

 first I had much difficulty in obtaining in every analysis the same 

 results by the new volumetric and Pettenkofer's methods ; at last I 

 found that the only cause of error left was the displacement of the 

 surface of the oil, occasioned by working the bell -jars too fast ; when 

 moved slowly, the results by both methods invariably agreed. I con- 

 clude that air containing about 4 per cent, of carbonic acid will yield 

 within 0*5 per cent, of the proportion which would be obtained by 

 Pettenkofer's method, and air containing 1 per cent, of carbonic acid 

 will give a result within 1 per cent, of that to be obtained by Petten- 

 kofer's. In the last analysis I have quoted, air containing less than 

 1 per cent., or 0*9327 per cent, only of carbonic acid, gave with the 

 two methods a difference of \ 0*45 per cent., a very satisfactory 

 result. 



In conclusion I must beg to return my best thanks to Professor 

 Schafer for having kindly allowed me the exclusive use of a room in 

 the Physiological Department of University College where these 

 researches were carried out; to Mr. W. Parkinson, of the firm of 

 Messrs. William Parkinson and Co., gas engineers, who constructed 



