38 JOURNAL AND PROCEEDINGS. 
by the methods of the mathematical theory of probabilities (whose 
success in solving the practical problems of life insurance may be 
cited as evidence of its reliability), has led to the conclusion that the 
molecules in one cubic inch of air, if spread out side by side ina 
single layer, would cover a surface of thirty-five square yards, 
In the meanwhile a careful repetition of Boyle’s experiments on 
the compression of gases, had shown that his “law” was to be re- 
garded merely as an approximation to the truth, neither air nor any 
other gas behaving strictly according to its requirements. ‘This 
discovery which seemed at first sight to remove the whole found- 
ation from the Kinetic Theory, directed attention to the fact that in 
the earlier calculations of Bernoulli no note had been taken of 
the space filled by the gas molecules themselves, and it was found 
that the deviation of air from the strict letter of the law of Boyle 
could be accounted for by assuming that the molecules in three 
thousand cubic inches of air actually occupy one cubic inch, the two 
thousand nine-hundred and ninty-nine cubic inches being the 
‘‘empty space” whose quantity varies inversely with the pressure. 
Thus by adding a little to the precision of the theory, its annihilation 
by experiment was avoided, or as its enthusiastic champions put it 
facts which.seemed at first te threaten the very existence of the 
theory, proved in the end but a means of penetrating still further 
into the secrets of the molecular world. 
Once given the actual volume of the molecules in one ounce of 
air and the maximum area they can be spread over, a simple process 
of division is all that is necessary to find the diameter of each in- 
dividual and another to arrive at their total number and the weight 
of each. The diameters and weights turned out, as might be expected, 
very small—17 x 10° 1 inches and 10°” grains respectively ; the 
total number very large about 3 x 10”. 
In this short sketch I have purposely avoided attempting any- 
thing like a full account of the modifications and applications of the 
theory (such for example as its prediction of the specific heat of 
mercury vapour), my wish being merely to give some idea of the 
means by which the various ‘‘ molecular dimensions” have been 
arrived at. Some general information on this point is important 
because there has always existed a large class of admirers of this 
Kinetic Hypothesis, who not knowing how these values had been 
