404 Mr. J. Stevenson on the Chemical and 



must be a truly wonderful degree of exactness in the adjust- 

 ment of the amount of oxygen on the earth to the amount of 

 oxidizable substances, if they balance each other down to the 

 1000 000 part of the total weight of either. Looked at from a 

 purely abstract standpoint, the standpoint of mathematical pro- 

 bability, so very exact a decree of balancing is extremely impro- 

 bable. We know practically nothing of the principles on which 

 the various chemical elements are distributed on the earth and 

 throughout the universe ; but let us assume for the moment 

 that there is no principle at all, or nothing more definite than 

 chance. On this principle, or want of principle, the proba- 

 bility that the total amount of oxygen on the earth will 

 amount to any particular figure is substantially equal to the 

 probability that it will amount to any other figure, out of a 

 practically infinite number of figures, which are limited only 

 by the condition that they must be considerably less 

 than the total weight of the earth itself — the limit being some- 

 thing like one-half or one- fifth of the weight of the earth. 

 As the possible figures, all equally probable, are practically 

 infinite in number, the chances against any one figure being 

 the correct one are practically infinite. If, then, we do not 

 know, even on the roughest approximation, the amount of 

 oxygen on the earth, but have two theories regarding a 

 particular terrestrial problem, one of which requires that the 

 total oxygen on the earth should be exactly, or almost exactly, 

 of a particular amount, viz., just a very little more than the 

 amount that is equivalent to the total oxidizable matter on the 

 earth ; while the second theory admits of the amount of oxygen 

 being anything at all within a very wide range — from a rough 

 (but inadequate) equivalence to the total oxidizable matter 

 down to a very small fraction — say the 15V0 P ar ^ °^ ^ na ^ 

 amount, then, obviously, the latter theory has on a priori 

 grounds an incomparably greater probability of being the 

 correct one. Clearly, therefore, the inference from the above 

 line of reasoning is that there is a deficiency of oxygen on 

 the earth relatively to the total amount of oxidizable matter. 

 The objection may be made that certain other elements, viz., 

 chlorine, bromine, iodine, and fluorine — which have not much 

 affinity for oxygen but which have a strong affinity for other 

 elements, and may therefore be regarded as rivals to oxygen 

 or substitutes for it — may possibly be present on the earth in 

 large quantity. The reply to this is a simple one, viz., that 

 they are met with to only a small extent (relatively speaking) 

 on the surface of the earth ; and even if they were present in 

 much larger quantities deeper down, that circumstance would 

 not materially affect the force of the above line of reasoning. 



