260 
MR. J. J. MANLEY ON THE APPARENT CHANGE IN WEIGHT, ETC. 
From the above tabulated results it may be seen that there were 72 5-point 
determinations of the relative weights of the reaction vessel and its counterpoise 
before effecting the chemical reaction, and 63 similar determinations after. The 
maximum variations observed amongst the several daily mean values (Column 3) 
were 0'63 before the reaction and 0'23 after; these variations in the R.P. correspond 
respectively to '014 and ’005 mgr. The mean values of all the daily average 
differences in the R.P. (Columns 5 and 7) before the reaction are +0’20 and —0T4 
respectively, or a final mean variation +0T7. After the reaction the corresponding 
values are respectively -l-0’35 and —0‘35, or a final mean variation = ±0'35. 
Expressing the two final means in the form of weight variations, we have for the 
first ± ‘004 mgr. and for the second + ‘008 mgr. The increase in the average 
variation here shown, though in itself small, is, we believe, definite, and we think 
that it is quite possible that as a result of the chemical reaction some at present 
unknown minute disturbing factor has been introduced; further research may bring 
this to light. 
Taking the final mean variation in the relative weights = ±'006 mgr., it appears 
that the additional refinements and precautions which we have introduced and 
described have resulted in raising the attainable degree of accuracy in weighing from 
the lower limit of ±'03 mgr. given by Landolt to one having just l/5th of that 
value. 
The weight of the reaction vessel and its contents may in round numbers be taken 
= 109 gr.; if, therefore, the difference Mg —be taken = ’001 mgr., instead of the 
smaller difference "0007 mgr. as found experimentally, the apparent increase in the 
total mass during the chemical reaction is of the order of 1 in 109,000,000 parts; but 
this, it may be observed, is 6 times smaller than the average variations exhibited 
throughout these final experiments. 
Reviewing then the whole available evidence, we are led to conclude that this 
present research has tended but to confirm the truth of an almost universally 
accepted belief, that a given total mass is an unchanging and unchangeable quantity. 
In conclusion, I desire to acknowledge my indebtedness to the Government Grant 
Committee of the Royal Society for the continued loan of certain apparatus which 
has considerably facilitated this research. My best thanks are due to Mr. R. T. 
Gunther, Fellow of Magdalen College, and to Mr. N. V. Sidgwick, Fellow of 
Lincoln College, Oxford, for the valuable assistance which both have rendered me. 
Lastly, I would express my appreciation of the kindness shown me by the Radcliffe 
Observer. In the absence of data afforded by Dr. Rambaut’s continuous barograph 
records it would have been difficult to apply certain minute but necessary corrections 
to the final R.P. values of the barium chloride and sodium sulphate experiments. 
