SUMMARY 19 



Thus it is not betrayed by any kind of discrepancy in 

 applying the usual methods of measurement. The dis- 

 crepancy is revealed when we change the standard 

 motion of the measuring appliances, e.g. when we com- 

 pare lengths and distances as measured by terrestrial 

 observers with those which would be measured by 

 observers on a planet with different velocity. Provision- 

 ally we shall call the measured lengths which contain 

 this discrepancy "fictitious lengths". 



According to the Newtonian scheme length is definite 

 and unique; and each observer should apply corrections 

 (dependent on his motion) to reduce his fictitious lengths 

 to the unique Newtonian length. But to this there are 

 two objections. The corrections to reduce to Newtonian 

 length are indeterminate; we know the corrections 

 necessary to reduce our own fictitious lengths to those 

 measured by an observer with any other prescribed 

 motion, but there is no criterion for deciding which 

 system is the one intended in the Newtonian scheme. 

 Secondly, the whole of present-day physics has been 

 based on lengths measured by terrestrial observers 

 without this correction, so that whilst its assertions 

 ostensibly refer to Newtonian lengths they have actually 

 been proved for fictitious lengths. 



The FitzGerald contraction may seem a little thing 

 to bring the whole structure of classical physics tumbling 

 down. But few indeed are the experiments contributing 

 to our scientific knowledge which would not be invali- 

 dated if our methods of measuring lengths were funda- 

 mentally unsound. We now find that there is no 

 guarantee that they are not subject to a systematic kind 

 of error. Worse still we do not know if the error 

 occurs or not, and there is every reason to presume 

 that it is impossible to know. 



