CHANGE IN WEIGHT DURING CHEMICAE REACTION. 
249 
From tlie results obtained during the wiping of a cliarged flask {vide supra), it was 
inferred that the less frequently we repeated the operation of wiping the reaction 
vessels, the greater would be the reliability of final results drawn from a whole series of 
experiments. Again, the oftener the balance case is opened the greater will be the 
risk incurred from falling dust, varying temperature, and moisture effects ; and with 
oft-repeated reversals we incur the further danger due to possiljle minute alterations in 
weight brought about by the friction between the hooks of the reaction vessels and 
those of the balance stirrups. 
Further, we were convhiced that the elaborate precautions taken Ijy Landolt for 
centering the masses under comparison upon the balance pan were unnecessary. 
Modern precision balances are so constructed that masses placed in their pans or 
suspended from the hooks above are automatically centred; but in order to remove, 
as far as our own balance was concerned, any doubt that might arise on this point, a 
number of experiments carried out with the aid of the inverted U-tubes conclusively 
showed that the contained liquid might be either equally or otherwise divided between 
the two limbs of the vessel without producing any perceptible change in the apparent 
weight of the whole. 
After the two Jena glass reaction vessels, henceforth termed X and Y, had been 
similarly wiped, they were suspended from the hooks of the balance stirrups and 
equipoised with the aid of a small weight and the rider; the balance pans and their 
contents were next completely enclosed by the copper cylinders (p. 237), and tlie tubes 
delivering dry and purified air, placed in situ. The shutter of the balance case having been 
closed the beam was released and fatigued for a prolonged period; the actual weighing 
was then commenced and continued at intervals during successive days. For every 
determination of the relative weights, five pointer readings were taken for deducing 
the R.P. and the temperature of the balance beam noted. By means of the tempe¬ 
rature coefficient of the balance, the various B.Ps. were then reduced for a standard 
temperature of 16° C. The streams of purified air were always arrested during the 
time of the actual weighing, but allowed to flow freely at all other times. 
From September 28 to October 2, preliminary weighings only were undertaken; for 
it follows from the theoretical considerations which have already been given that it 
would be quite useless to attempt thus early to discover the true relative weights of 
the two vessels and tlieir contents. The results obtained during the preliminary 
weighings served chiefly to indicate the cessation of temperature fluctuations within 
the contents of the vessels which had, of necessity, been so freely handled. As may 
be seen from Table V., p. 250, in which the summarised results are set forth, the 
fluctuations as represented by the variations in the R.P. gradually die away and 
ultimately become quite small. 
From the above evidence it was concluded that the vessels and their contents had, on 
October 2, assumed an almost truly normal state; but it was thought to be safer to 
wait yet another day before beginning a final comparison of the weights of the two. 
VOL. CCXII.^—A. 2 K 
