168 Transactions. — Miscellaneous. 



Moreover, just as wliere the ships are most thickly spread we should 

 find most wreckage, so in the thickly strewn galaxy we find nearly all the 

 phenomena which tangential collision may be shown to be capable of pro- 

 ducing. 



But even without the influence of gravitation, the indiscriminate motion, 

 the vast size, and countless number of the stars, must result in occasional 

 collisions. Let us examine the case of the molecules of a gas as an analogy. 



The velocity of a gaseous molecule at ordinary temperature is nothing 

 to the stellar velocity, and the space occupied by the molecules in a good 

 vacuum is many million times less than the volume of the molecules, yet 

 spectrum analysis shows that such molecules are in incessant impact, and 

 there is practically no gravitation. The explosion of a large vessel of oxygen 

 and hydrogen appears instantaneous, yet each molecule of hydrogen and 

 each of oxygen must find the other in their excursions during the duration 

 of the explosion. Newcombe compares the motion of the stars with the 

 dance of the molecules, and surely if the impacts are so inconceivable in 

 number in the one case, they may sometimes happen in the other. As 

 though to make the analogy stronger, the case of chemical mutual exchange 

 may be considered a kind of molecular partial impact. 



To repeat an argument I have used before, cosmical impact is actually 

 in constant occurrence. Every meteorite that strikes the earth is an 

 instance of it, and what occurs on a small scale must surely be expected on 

 a larger, but is of course of rarer occurrence. But supposing our present 

 knowledge, instead of rendering it highly probable (as it is thus seen that it 

 does) that partial impact occurs, made it instead appear improbable, then 

 the vast number of celestial phenomena which receive a perfectly scientific 

 explanation by partial impact — and are so utterly inexplicable on any other 

 theory that positively no explanation exists — furnish such remarkable 

 evidence of these impacts that their apparent improbability would have to 

 be set down to our imperfect knowledge rather than to any improbability in 

 fact. 



Thus it is apparent that though Croll's idea of a direct impact may be 

 highly improbable, and if it occurred could not possibly explain his diffi- 

 culty, yet tangential collision — whose phenomena have been described under 

 the title of " partial impact" — is both in itself probable, and is known to be 

 constantly occurring on the small scale. Therefore, when this considera- 

 tion is combined with the amazing mass of evidence the phenomena of the 

 heavens offer, the probability of such impact is erected into a position as 

 near certainty as it is probable that human knowledge can attain. But if 

 Proctor means that all impacts are improbable, he does not seem to me to 

 be logically consistent, for his idea of the Universe is that it is eternal, 



