THE UNFATHOMED UNIVERSE 23 



than approximate, extension or intensification of inquiry 

 may show that it does not apply beyond certain limits. Thus 

 the law of gravitation, which must be very near to perfect 

 accuracy when applied to planetary distances, may not hold 

 either for very minute molecular distances or for immense 

 stellar distances. 



(3) Even generalisations that work well and must bear 

 a close correspondence to reality, since they afford a basis 

 for effective prophecy, may require some modification, in 

 their setting at least, in the light of some new fact or idea, 

 of great magnitude. Thus Prof. Frederick Soddy writes: 

 " It sounds incredible, but nevertheless it is true, that 

 science up to the close of the nineteenth century had no sus- 

 picion even of the existence of the original sources of natural 

 energy. . . . The vista which has been opened up by 

 these new discoveries [of the radio-active properties of 

 some substances] is without parallel in the whole his- 

 tory of science" (Harper's Magazine, December, 1909, p. 

 53). 



(4) We cannot pass over the caution suggested by the 

 Michelson-Morley experiment, which showed that scientific 

 observations cannot transcend the system in which they are 

 immersed. In Prof. Wildon Carr's words, " It showed us v - 

 that observers within a system of reference, in uniform move- 

 ment of translation relatively to other systems, have no 

 absolute standard by which they can determine their move- 

 ment. There is no absolute ether, no absolute space, and no 

 absolute time, by reference to which we can determine our 

 movements" (1918, p. 21). And Prof. Max Planck 

 writes of this new idea of the relativity of time : " With 

 the revolution which it brings about in our conception of 

 the physical universe, no other is comparable, in range 



