In the course of routine bacteriological examinations of milk carried out at one of the United Dairies pasteurising depots during the past three years, a remarkable lack of constant relationship between the numbers of colonies reported on the 1/10, 1/100 and 1/1000 dilution plates, when standard agar (the Ministry of Health formula) was used, has been observed. When, however, standard agar, enriched with 0·5 per cent, milk, has been used as a plating medium, many of the discrepancies between different dilutions of the same milk have disappeared.