T. HARPER'S METAPHYSICS OF THE SCHOOL, in. i. 585 teresting and instructive, if there were space, to compare the animadversions of these two critics. They both agree in this, that ' ' the question touching the ultimate constitution of bodies is beyond the sphere of physics," and " must be left to the de- cision of the Supreme Science," though they are not at one as to what that is in prime assumption and method ; and yet they are both nearly willing (Father Harper almost if not quite so, p. 332) to let physicists use their atomic theory meantime as an explaining expedient, provided they do not hold it to be " sub- stantially true ". Yet even the latter is not always quite assured of the serviceableness of this " ingenious " and " purely ideal " device, in fact many times half withdraws his permission, as on pp. 335-341, where he remarks on the lack of " direct physical evidence " in favour of the provisional hypothesis, the doubts of distinguished chemists as to its truth and use (quoting Cooke and Miller), the anomalies and inconveniences it has led to in the modern basis of chemistry ; and concludes that the science might very well have come, and would remain " unimpaired," without the theory in question. In this context what he says about the law of Avogadro specially deserves attention. Un- like the laws of Mariotte and Boyle, it presupposes the truth of the molecular theory, though it is based on a fact that is inde- pendent of it or any other hypothesis. There are facts which he very well shows are irreconcilable with it ; for, ia sum- mary, " the quantivalence of one and the same element changes . . . even relative to one and the same component ". And yet this law of Ampere or Avogadro holds, according to Prof. Cooke, " the same place in chemistry that the law of gravitation does in astronomy " ! Again, to the Aberglaube of the atomists our author attributes an explanation of isomerism entirely un- satisfactory. Confess your ignorance and wait, he says, rather than accept such very artificial magic. Be content to refer the symbols and ratios at present in use to sensible quanta and not to " supersensuous " units, which the metaphysical atoms and mole- cular assemblages of atoms are ; and you will find them quite as intelligible and valid, nay more so. "Affinity, proportion, and the relation of matter (mass?) to volume," the three foundations of chemistry, will still stand fast and sure. In these pages the two questions of the truth of atomism and its validity are mostly inter- woven, and very naturally so, as what is not true can scarcely be useful, nor what is useless be true ; and in this parallel treatment one may be willing to go with F. Harper without serious risk of confusion. This applies to the two pages of " Vacuum " and " Actio in distans," by no means all vacuous or far-fetched pages, in which he has used the very finest edge of his dividing mind, and which come into his criticism of atomism thus : The atomic theory in modern chemistry presupposes vacuum and its corollary, action at a distance, " in order at once to defend the isolation of molecules and atoms, and the existence of chemical action".
