CHANGE IN WEIGHT DURING CHEMICAL REACTION. 
243 
odour frequently manifested itself when solid caustic soda was used ; and it was 
suspected that the odour indicated the presence in the air of minute particles of the 
more or less hydrated reagent. Our suspicions were tested in the following way :— 
Two clean porcelain dishes, each 4^ inches in diameter, were taken ; the one was 
charged with short sticks of caustic soda, and the other with a little “ conductivity ” 
water which was then tinted with methyl orange; the two dishes were placed side 
by side and covered with a large bell-jar. After the lapse of several hours the bell- 
jar was removed and held in a horizontal position ; a Bunsen flame was then 
introduced ; the flame was at once tinged faintly but perceptibly yellow, thus showing 
the presence of soda particles. It was next observed that the colour of the methyl 
orange was markedly yellow, thus indicating the presence of free alkali ; we therefore 
proceeded to titrate the alkaline solution with N/lOO HCl, and found that between 
4 c.c. and 5 c.c. of the acid were required for the neutralisation of the absorbed alkali; 
from this it follows that the amount of caustic soda present in the water was equal to 
nearly 0'2 mgr.* Acid was then added until the colour of the indicator was 
decidedly pink and the dishes were again covered with the bell-jar. After some 
considerable time it was found that the excess of acid had been neutralised, the methyl 
orange having re-assumed a bright yellow tint. From these observations we conclude 
that the moisture in the air is so powerfully attracted by the drying reagent tliat a 
perceptible rebound from the absorbing surface follows. Although no experiments were 
made, it appears reasonable to conclude that similar phenomena would attend the use 
of other desiccating bodies. 
In concluding this brief criticism of existing methods, we desire to remark upon 
some possible inaccuracies that may follow the use of sulphuric acid as a desiccating 
reagent. 
On exposure to air concentrated acid quickl}^ becomes diluted over its surface ; as 
dilution proceeds the acid develops an apprecia])le vapour tension; it therefore 
follows that surfaces exposed in the vicinity of the acid will first lose their water 
skins and so become lighter; then, with prolonged exposure, they will acquire acid 
skins. I myself have frequently noted an increase in the weight of a porcelain 
crucible over-dried (if I may use the term) in this way.t 
Seeing then that the usual method for drying the air can but lead in the direction 
of uncertain values it was abandoned for the following plan which proved both 
convenient and highly satisfactory. 
* This experiment has recently been repeated. The volume of N/lOO acid required for neutralising 
the alkali absorbed by the water in the dish = 1'5 c.c., or 1/3 the amount required in the experiment 
described above. 
t The following evidence tends to confirm the accirracy of the views expressed above. The upper 
portions of the interiors of 18 Scheibler desiccators, which had been in constant use for 8 weeks (some 
being used much more frequently than others), were examined by wiping them with wet blue litmus 
paper; in 12 cases a distinct acid reaction was obtained. 
2 T 2 
