228 
MR. J. J. MANLEY ON THE APPARENT 
Before proceeding to deal with my own investigations in this held of research, it may 
not be inappropriate first very briefly to recall the chief features of Landolt’s work 
and conclusions. 
From Landolt’s memoir we find that in all some 15 different chemical reactions 
were used; these were, for very obvious reasons, brought about within hermetically- 
sealed vessels. Convenient methods were adopted for ensuring a very slow mixing of 
the reacting bodies, the period allowed for this purpose varying from two to four days, 
and therefore the heat evolved during the chemical transformation was so dissipated 
that no very perceptible rise in the temperature of the vessels and their contents 
occurred. By following this plan it was hoped that all measurable temporary changes 
in the volumes of the reaction vessels would be avoided. 
In connection with the weighings, devices were introduced for securing as nearly as 
possible a truly symmetricpJ disposition of the combined masses of the suspended 
vessels, their contents and carriers, about the vertical lines passing through the 
terminal knife-edges of the balance. The relative weights of the reaction vessel 
and its compensating counterpoise were determined according to the method 
advocated by Gauss. The oscillations of tlie balance beam were observed with the 
aid of a telescope placed at a distance of 3 metres from the balance; and from the 
same distance (3 metres), by means of suitable apparatus, the balance was both loaded 
and the positions of the contents of the pans interchanged. Landolt effected most of 
his final weighings at a temperature approximating 19° C.; this temperature was, for 
convenience, selected as a standard one, and if at the time of weighing the 
temperature deviated from the standard value, then from a knowledge of the 
temperature coefficient of the balance it was possible to calculate the value which the 
observed resting-point of the beam would possess at 19° C. 
It appears that the balance prior to the majority of the weighings was duly 
fatigued, but whether the necessary fatigue was designedly imposed or was merely 
the outcome of accidental convenience it is difficult to say. 
The fully corrected final values derived from the whole of Landolt’s most 
trustworthy observations are tabulated on pp. 153-4 of his memoir. From those 
values we find the minimum apparent change in mass during any one chemical 
reaction to be ±‘001 mgr., and the two maxima -{-■021 mgr. and —’038 mgr. 
respectively. Of the 48 final values, 23 show an apparent increase in weight and the 
remaining 25 an apparent decrease. The mean positive difference deduced from all 
the tabulated results we find to be = '009 mgr., and the mean negative clifierence 
= ‘015 mgr. 
Combining the whole 48 corrected values, we obtain as a final result an apparent 
decrease in weight = "006 mgr. 
Now Landolt estimated his possible experimental error to be approximately = ± ‘03 
mgr.; in his memoir he also expresses the opinion that ± '03 mgr. represents the 
present attainable degree of accuracy in weighmg ; it will therefore be observed that 
