This might make interesting reading:
Can any of the illuminati offer a way of distinguishing between a real drift in physical constants and simple evolution of experimental technique? The cyclic nature of the value of c that Sheldrake mentions on this page, for example, seems rather hard to explain, but does it need explaining at all? One thought occurs to me; if c varies across finite volumes of space, then you'd see refraction effects there...if the volume is big enough or the difference in c is great enough, then maybe you'd be able to detect significant effects..."gravitational" lensing without intervening galaxies?
His research on ESP in dogs is more fun. Pretty convincing for those of us who've had the experience with our own pets.