462 SCIENCE PROGRESS 



words : — The laws of nature are independent of the state of 

 motion of the system of reference provided this is unaccelerated. 

 This means in more particular terms, that if two sets of ob- 

 servers were stationed in two different worlds, and each set 

 made ideally careful and precise measurements on all natural 

 phenomena, then they would each summarise their observations 

 in precisely the same laws, provided each of the worlds was 

 moving uniformly with respect to the other. So it would 

 appear that natural phenomena would fail to decide as to which 

 of the two worlds was absolutely at rest in space and which in 

 motion. Such phenomena could only yield a measure of their 

 relative motion. Indeed, as far as purely mechanical pheno- 

 mena are concerned the principle is as old as Galileo, and is 

 employed by every student who works out a problem in Rigid 

 Dynamics involving the use of " Moving Axes." It was the 

 continued failure of optical phenomena to reveal any difference 

 in the measured velocity of light, whether the light was travel- 

 ling in the direction of the Earth's motion, or against it, or 

 across it, which led Einstein to extend the principle, so as to 

 include all phenomena, optical and electromagnetic, and in 

 particular to enunciate the special case of relativity known as 

 the principle of the constancy of the velocity of light, viz. that 

 the velocities of light measured with respect to two different 

 systems of reference will be the same if the two systems are ii 

 uniform, relative motion one to the other. It follows from this 

 postulate that in transforming from a set of axes fixed in one 

 system of reference to a set of axes fixed in the other, not onb 

 are the co-ordinates of a point with respect to the one set line; 

 functions of the co-ordinates referred to the other set and the 

 time (which is to be expected), but the time of the occurrence 

 of an event at this point in the first system of reference is also 

 linear function of the co-ordinates of the point in the seconc 

 system and the time of the event in the second system. It 

 this second fact which constitutes such a novel departure froi 

 our ordinary notions as regards the invariability of the time 

 measure of an occurrence. Further it appears that even the 

 linear relations connecting the co-ordinates in one set with the 

 co-ordinates and time in the second set, are not those derivee 

 from the usual " principle of moving axes " as employed in 

 works on Dynamics. In fact these relations are so modifiec 

 that they have precisely the same form, whether one is trans- 



