ON THE USE OF THE HYDROMETER OF TOTAL IMMERSION. 261 



sample can be avoided by rinsing the bottles well in heated 

 distilled water for several days beforehand. Contamination of 

 the sample on opening the bottle can also with care be easily 

 avoided. The only source of error which cannot altogether be 

 overcome is the evaporation of the water or condensation of the 

 moisture of the air, as the temperature of the sample is respec- 

 tively higher or lower than the dew-point of the surrounding 

 atmosphere. ^ 



Nor is the error thus occasioned constant. The rapidity 

 with which the water evaporates is dependent upon the temper- 

 ature and amount of moisture in the air, and also upon the in- 

 side diameter of the cylinder. The increase in the specific 

 gravity of the water sample, produced by evaporation, will finally 

 be greater in a small sample than in a large, when otherwise 

 the circumstances are similar. 



It appears also, in the determinations that I have made, 

 that the second determination of a sample nearly always shows 

 a rather higher specific gravity than the first. There is exclu- 

 sively question of evaporation here, as the temperature of the 

 sample at the time of determination has always been higher 

 than the dew-point of the air in the room, and the amount of 

 moisture in the atmosphere in a room in Kristiania in the winter 

 is particularly small. 



In order to obtain some idea of the rate of evaporation 

 under ordinary conditions in the laboratory, I determined the 

 specific gravity of a sample with the hydrometer of total immer- 

 sion, and after letting it stand uncovered for some time, made 

 a fresh determination. The following table gives the result of 

 two such experiments. 



1 All determinations were made with the cylinder uncovered. 



