SUPERLUMINAL LIGHT PROPAGATION. Scientists at the NEC Research Institute in Princeton have performed an experiment in which the group velocity of a light pulse traveling through a special medium appears to be faster than c, the speed of light in a vacuum, without, however, violating the principle of causality or the theory of relativity. Such experiments have been performed before and have exploited the fact that a finite pulse of light is necessarily the sum of an ensemble of waves at different frequencies. One therefore speaks of a "phase velocity" for component waves and the "group velocity" for the pulse as a whole.

When such an ensemble enters a medium with a frequency-dependent index of refraction, interesting things start to happen. In a Harvard experiment last year, for example, the component light waves of a pulse passing through a Bose-Einstein condensate were affected in such a way as to yield a group velocity of only 17 m/sec (Hau et al., Nature, 18 February 1999).

Working in the other direction, manipulating the component waves in order to achieve a higher group velocity, is more difficult to establish since it usually occurs when the index of refraction is varying rapidly in a frequency range where the light is being absorbed by the medium; hence the light pulse can be severely distorted or attenuated, making it difficult to detect superluminal effects.

In the NEC experiment, by contrast, the medium in question, a cell filled with a gas of cesium atoms, does not absorb light at the crucial frequencies but actually enhances the light through a type of laser action; that is, the cesium atoms are promoted into an excited state and contribute to the light pulse when it travels through. Consequently the pulse shape is largely preserved even as the component waves interfere (through a process called anomalous dispersion) in such a way as to shift the pulse forward in time by a tiny amount, about 1.7% of the original pulse width, compared to the situation in which the cell is not present. According to the NEC researchers, "the peak of the pulse appears to leave the cell before entering it." This superluminal behavior does not contradict the principles of Einstein's relativity theory, but it might well encourage further discussion among scientists about how exactly to specify the onset of light signals. (Wang et al., Nature, 20 July /pnu/2000/.)