More than one correspondent has brought against NIAT the counter-example of the velocity of light, held to be not merely an absolute but a scientific one. IC has responded by pointing out that this uses ‘absolute’ in a different sense from NIAT, a more limited sense which amounts to hardly more than ‘invariant.’ It now appears that the velocity of light may not be absolute even in that weak sense. Consider this:
In 1859 Hippolyte Fizeau published, in Annales de chimie et de physique, a fine experiment which ‘showed that the velocity of light increases in a medium according to the formula: v (1 – 1/n2) where v is the velocity of the medium, and n its refractive index.’ 
If the velocity of light varies with the medium, then the claim even for its invariance does not stand up. Would any reader enjoying familiarity with these matters care to comment?
 C. G. Gillispie 1960, The Mechanization of the World Picture, Princeton etc.: OUP, 427.
from Ideological Commentary 63, February 1994.