230 
MR. J. J. MANLEY ON THE APPARENT 
secured in positions which fall within the outer struts of the beam. The practically 
perfect rigidity of such a beam is proved by the fact that the sensibility of the 
instrument remains, so far as can be demonstrated l)y actual experiment, quite constant 
for all loads varying from 0-200 gr. The knife-edge blocks and the planes they 
engage are of agate ; beam, pans and stirrups are highly gilt. The beam is enclosed 
by one of the author’s protecting inner cases which eftectively excludes all disturbances 
which would otherwise be produced within the beam during the time the ordinary 
case is open for the loading and unloading of the pans, &c. By means of a platinum 
wire bolometer placed within the beam case, we were at any time able to test the air 
temperature in the immediate neighbourhood of the extremities of the beam ; the 
bolometer was capable of detecting temperature differences so small as l/l0,000° C. 
The actual temperatures of both the interior and exterior of the beam case were 
noted by means of delicate mercury-in-glass thermometers ; with these instruments 
we were able to read to l/l00° C. The sensibility, S, of the balance, was set 
moderately high ; it was equal to 45 scale divisions per 1 mgr. and was kept unchanged 
throughout. 
(3) The Reacting Substances. —In commencing this investigation some little 
difficulty was felt in selecting a reaction likely to be the most suitable and con¬ 
venient for the object in view. Of the 15 clieniical reactions chosen by the original 
investigator, a w'ell defined apparent decrease in the total mass of the reacting 
substances was usually obtained by means of ferrous sulphate and silver sulphate ; 
and this was particularly the case during the earlier experiments. In two other series 
of experiments, silver nitrate was substituted for silver sulphate, the second body being, 
as before, ferrous sulphate. For the one series the final result was -H'003 mgr., and 
for the other — ’003 mgr. ; the mean final value was therefore = 0. Our choice 
ultimately fell upon the latter reaction, which in Landolt’s hands had, from the 
standpoint of the present-day theory of the strict conservancy of the total masses of 
reacting bodies, led to the most satisfactory conclusions. 
Landolt states the reaction which takes place between silver nitrate and ferrous 
sulphate in the form of the equation 
6 AgN 03 + 6 FeS 0 , = 6Ag + 2 Fe 3 (S 0 ,) 3 +Fe,(N 03 )e. 
He, however, remarks, that the reaction is not complete. In the sequel it will be 
shown that this conclusion is correct. It will also be shown that had Landolt 
experimented more freely and extensively with silver nitrate and ferrous sulphate the 
degree of concordance observed by him would, in all probability, have been considerably 
modified. 
(4) The Reaction Vessels and their Preliminarg Treatment. —For the first series 
of experiments, the form chosen for the reaction vessels was that largely adopted by 
Landolt ; it is best described as an inverted U, the two limbs of which were 
cylindrical bulbs, each having a capacity of about 70 c.c. Two such vessels were 
