After indicating the basis of the technique, its use with a cryogenic SQUID magnetometer is outlined, particular emphasis being given to the need for stringent reliability criteria. Comparison of results for some samples is made with results obtained using the Shaw alternating‐field technique, and the agreement is satisfactory. Determinations made on 14 samples fired during the last 150 years give results which on average are in close agreement (within 2%) with observatory‐based values, the standard deviation of a sample from the predicted value being