142 THE THEORY OF GRAVITATION. 



the consequences of the theory of atomic collisions would have been 

 unequivocally in favor of the sole right interpretation (equal accelera- 

 tions in equal times). 



The union of the several branches of this conclusion forms not only 

 a philosophic truth of extreme interest, but one from which a very 

 useful consequence may be drawn, which is that in spite of the greater 

 weight due to a posteriori researches a priori ones are not to be wholly 

 neglected, since they may greatly accelerate the success of the former. 

 Already some impartial philosophers are agreed that such conjectures 

 if lucid and capable of evaluation might be useful to the most rigorous 

 physicists, were it only in suggesting to them definite points of view 

 from which to direct experiment, in the place of that indecision in which 

 the mere vague wish for new investigation has often left them. 



Let us clearly understand that such speculation is only permissible 

 for the sake of occupation when the skill and patience which new 

 observation and experiment require are lacking. We ought to be 

 thoroughly informed as to all previous observations and experiments 

 on the subject and to keep these steadily in view in forming hypotheses, 

 which are to be tested by them with the aid of every help that mathe- 

 matics can give in examining as to the exactness of their agreement. 



Finally, it is such an agreement rather than any elaboration of method 

 which brings conviction to most students of any physical theory, and 

 this whether they are aware of this agreement before their acquaintance 

 with these methods or whether a study of the method led them to the 

 agreement. 



If the disciples of Epicurus had been as fully persuaded of the 

 sphericity of the earth 1 as they were of its flatness, 2 then instead of 

 conceiving their atoms to move in nearly parallel paths, as was suited 

 to a directive force perpendicular to a plane surface, they would 

 undoubtedly have attributed to them motion normal to the surface of 

 a sphere, and consequently directed at all points toward its center. 3 

 An example of such a condition as I have in mind would be furnished 

 if it hailed simultaneously in all the countries of the earth. 



1 Plato and Aristotle had discoursed at great length upon the sphericity of the 

 earth; Archimedes and Aristarchus had assumed it; Thales and Zeno had taught it, 

 and all the astronomers believed it. (See the Tirnreus of Plato, the close of the 

 second book of Aristotle upon the Heavens, the Hour-Glass of Archimedes, and the 

 tenth chapter of the third book of Plutarch upon the Opinions of the Philosophers.) 



'-'Neither Epicurus nor Lucretius discovered the figure of the earth. But it seems 

 probable that they conformed to the opinions of Democritus upon all questions -where 

 they did not expressly oppose him. Moreover, Gassendi (in his Commentaries on 

 Epicurus, p. 213 of the edition of 1649) alleges strong reasons for believing that 

 they supposed the earth's surface to be flat. 



:; Instead of which they entirely rejected this centripetal tendency. 



