On the probability of detecting nebulae which act as gravitational lenses

Recently various authors (1,2) have again (3) considered the possibility of observing the image of a distant star A whose light is bent around some nearer star B. For reasons discussed by these authors, the probability that the mentioned effect will ever be observed with stars is vanishingly small. The general feeling therefore was that the idea of gravitational lenses affords "perfect tests of general relativity that are unavailable," as Professor H. N. Russell (2) puts it.