i;8 THE LIMITATIONS OF SCIENCE 



nomena. If we know the distance and the time, we 

 say the velocity of light in empty space is the quotient 

 of the two. Such observations are quite inadequate 

 to settle experimentally the question of the constancy 

 of V. Our measurements of the distances are crude 

 for such a purpose, and also the light must travel part 

 of the way through a material medium such as the air. 

 Neglecting such astronomical methods, we have left 

 only determinations of V made on the earth where 

 the path is necessarily through matter. Here, what we 

 are to call the constant V can be found only by extra- 

 polation. For example, we find that the ratio of the 

 velocity in air and in water is about four to three and 

 agrees with their refractive indices. We find also that 

 the velocity in different gases as they are reduced in 

 density tends to a common value, which is independent 

 of the kind of gas and approaches the value 3 x io 10 

 centimeters. If we take this extrapolated value of 

 3 x io 10 centimeters for the velocity of light in empty 

 space and assume the refractive index of space to be 

 one, then we may calculate back and by this method we 

 find the absolute refractive index of air at ordinary 

 pressure to be 1.0002. Such an experimental method 

 and such reasoning are highly unsatisfactory. In the 

 first place, we start with measurements of lengths and 

 time which are subject to experimental verification and 

 calculate V, from these we pass to a supposititious V 

 where only one factor, the time, is subject to experi- 



