SLOMAN1 has claimed to have found, using optical microscopy in conjunction with chemical analysis, that a beryllium oxide/beryllium eutectic forms at grain boundaries in vacuum-cast beryllium containing 0.3 per cent oxide by weight. By contrast, other workers2 have since been unable to detect such oxide, either in the grain boundaries or present as inclusions in metal of comparable purity. It is still generally agreed, however, that the brittleness of beryllium is probably associated with some impurity, and our current results support the original claims by Sloman.