This paper points out a cause of error in the evaluation of new drugs. An investigator who gave a drug (veriloid) to patients was at first enthusiastic and felt friendly toward the experiment; this attitude was reflected in the relief of patients' symptoms and the decrease in their blood pressure. Later, for a variety of reasons external to the experiment, the investigator's attitude became sceptical and unenthusiastic and he regarded the whole project as an onerous chore. At