STEWART and his associates have developed a method for detecting a haemagglutinating antigen for rubella virus1 which depends on two discoveries: first, that the calf sera ordinarily incorporated into media for the maintenance of cells infected with rubella virus contain an inhibitor for haemagglutination and, second, that the best agglutination was seen with blood cells from newly hatched chicks. In their original method BHK21 cells, which have been shown to produce the virus in high titre2, were infected with rubella virus and maintained in 2 per cent kaolin-treated foetal calf serum. Although this method provided a signal advance in the serology of rubella, in our laboratory production of haemagglutinin has been variable and, when found, the peak titres observed were only 1/32 per ml.