We returned to Cambridge in October, and Arthur threw himself vigorously into the new year’s work. Then I fell ill, and the hope we had indulged was disappointed. Worse, the course of the illness revealed a condition which demanded the most complete series of operations which a woman can endure. Not only the past hope, but all future hope, was annihilated.

It was during my convalescence that the most remarkable incident of my life took place.

I was in great pain one afternoon, and wished to see the doctor. The nurse went to the study to telephone for him.

“Nurse!” I said, as she returned, “don’t lie to me. He’s not gone to Royston; he’s got cancer, and is too upset to come.”

“Whatever next?” said the nurse.

“It’s right he can’t come, and I was going to tell you he had gone to Royston; but I never heard nothing about no cancer.”

This was true; she had not been told. But the next morning we heard that my “intuition” was correct.

As soon as I was well enough, we began our experiments again. My powers had returned, and in triple force.

Arthur explained my “intuition” as follows: “The doctor (when you last saw him) did not know consciously that he had cancer; but subconsciously Nature gave warning. You read this subconsciously, and it sprang into your consciousness when you read on the nurse’s face that he was ill.”

This, far-fetched as it may seem, at least avoids the shallow theories about “telepathy.”

From this time my powers constantly increased. I could read my husband’s thoughts from imperceptible movements of his face as easily as a trained deaf-mute can sometimes read the speech of a distant man from the movements of his lips.

Gradually as we worked, day by day, I found my grasp of detail ever fuller. It is not only that I could read emotions; I could tell whether he was thinking 3465822 or 3456822. In the year following my illness, we made 436 experiments of this kind, each extending over several hours; in all 9363, with only 122 failures, and these all, without exception, partial.

The year following, our experiments were extended to a reading of his dreams. In this I proved equally successful. My practice was to leave the room before he woke, write down the dream that he had dreamt, and await him at the breakfast table, where he would compare his record with mine.

Invariably they were identical, with this exception, that my record was always much fuller than his. He would nearly always, however, purport to remember the details supplied by me; but this detail has (I think) no real scientific value.

But what does it all matter, when I think of the horror impending?