452 THE POPULAR SCIENCE MONTHLY 
The real significance of recent developments is in the fact that they 
change—in a way revolutionize—some of our ideas of things. And 
here let me say that proved facts and proposed theories should not be 
confused. A theory is simply a working hypothesis, invented for the 
purpose of explaining facts, to be discarded when facts are discovered 
with which the theory is not in harmony. <A theory may explain many 
facts, it may be generally accepted, it may have survived for generations, 
and be false. The phlogiston theory, the corpuscular theory, are two 
examples. Shall we say that the theory of the indestructibility of 
matter, and of the conservation of energy, are two others ? 
The usual chemistry text-book would have us believe in the inde- 
structibility of matter because the chemist can change the form of mat- 
ter almost at will, and in all the chemical reactions there is no loss of 
weight. In replying to this argument I wish to make three points: 
1. The balance, notwithstanding the statement of text-books, com- 
pares weights and not masses, and it is only because weight is assumed 
to be proportional to mass that we say we determine mass by the balance. 
What we really compare is the gravitational force which the earth exerts 
on two masses, and we have no @ priori right to assume that this gravi- 
tational force is absolutely independent of the state or molecular ar- 
rangement of the attracted body. Why, for instance, should we expect 
an absolutely uniform field of force about a crystal when that same 
crystal will, if placed in a proper solution, continue to grow symmet- 
rically, and perhaps replace a broken-off corner before beginning its 
growth? 
It is conceivable that there should be a loss of weight in chemical 
reactions and yet no destruction of matter. It is possible that mass 
and weight are not strictly proportional. If J. J. Thomson were not 
disposed to question the equation w==m.g. he would not have experl- 
mented with a pendulum of radium, and he would not now be experi- 
menting with a pendulum of uranium oxide. 
2. In the second place there is an apparent change of weight in 
chemical reactions as has been shown by several experimenters—notably 
by Landolt,® who found a loss in forty-two out of fifty-four cases. The 
chemical reactions were brought about in sealed glass tubes which gen- 
erally weighed less after the reactions than they weighed before. Later* 
:+ was found that some of these losses might be attributed to tempera- 
ture and volume changes. Whatever the testimony of the balance may 
have been, some of the reactions must have been accompanied by a loss 
of weight, for it has been proved by chemical means that such reactions 
are frequently attended by the escape of something through the walls 
of the glass tubes.® This loss is readily explained by the disintegration 
’Landolt, Preuss. Akad. Wiss. Berlin, Sitz. Ber., 8, pp. 266-298, 1906. 
‘Landolt, Preuss. Akad. Wiss. Berlin, Site. Ber., 16, pp. 354-387, 1908. 
5C, Zenghelis, Zeitschr. Phys. Chem., 65, 3, pp. 341-358, January 5, 1909. 
