"But I wasn't superstitious. No, I wasn't! At least, I didn't have what you'd call the superstitious feeling. In a twisted logical way, I was trying to be empirical. I judged everything by results, step by step.
"And of course, when I say empirical, I don't mean experimental. I didn't dare to omit any charms as a test, because I was too afraid of something going wrong. You can't be experimental about someone you love, any more than a sane sociologist would induce fascism in his own country just to study the consequences."
He listened, his mind working like a telephone exchange during an emergency, catching a thousand hidden connections in trivial memories of his life and Tansy's. Whenever she paused, he had a question to hammer at her.
"Those mirror decorations on your dresses, and belts, and handbags—why?"
She nodded, tiredly. "Yes, that's it. To ward off evil spirits, by reflecting them. I got the idea from Tibetan devil masks. Yes, and the reason I always use hooks and eyes is—you've guessed—to catch evil spirits that try to get at me."
The room was dark by the time Tansy had finished the catalogue of her activities at Hempnell, and said, "So you see, I've never been sure. I've wanted to stop, but there's been too much at stake. It always worked—or did as soon as I made the proper corrections in the charms. I've wanted to tell you, but I've never dared. Now that you know, I'm glad."
The voice was very tired now.
Then, in the darkness he began his argument. It was the old, old argument of science against primitive ignorance, but he brought it home to her with careful persuasiveness. The argument which starts out with a demonstration that superstition is only mistaken empiricism, and ends with psychoanalysis.
"Didn't it really begin much earlier than Hempnell, Tansy?" he asked at one point. "I mean, the seeds of it?"