Then, two days later still, on Friday, July 2nd, they had unexpectedly turned up together at my house. Now a definite note in the diary, written as a matter of fact in my own house (for he kept it instantly up to date), told me that on that day, July 2nd, he had "felt twenty-nine." True, he had later admitted the vagueness of these mere "feelings" as an index to age, but there it was for what it was worth, and it agreed with the impression I had myself formed, based on his vivid and ecstatic and momentary moods. Except when I had compelled him to speak of his book, Saturday had been the counterpart of Friday. That is to say, that during the whole of Friday and Saturday he had remained twenty-nine. Therefore (and omitting the loss of the years forty-five to thirty-five, now untraceable), during the twenty-five days from June 8th to July 3rd he had dropped a total of six years. So far so good; but that was not quite what I wanted to know. What I was trying to ascertain was a far more important thing--the shortest _actual_ time in which he had lost the great length of _apparent_ time. It would make the greatest practical difference in the world whether this figure were a high or a low one. And now groan, as I groaned, when you look at the four days between June 29th and July 3rd--those four days in which, in order that he might be at the very top of his power for the writing of his book, he had vehemently _denied_ his age, had juggled with it, wrestled with it, refused it, ignored it, vowed that a false middle was or should be a true one, and had hung as it were to a strap while the whole momentum of his being had tried to sway him in another direction. The entry for those four days was a mere question-mark with an open choice. It read: "Thirty-three--thirty?" And yet on the fifth day he had been twenty-nine! Now let us take the queried figures separately and subtract. If on the fourth day he had been the lower figure--thirty--then he had only dropped a year in a night. But if on the fourth day he had been thirty-three, then he had dropped four whole years in the same time. Either was possible, and yet in the one case the ratio was, appallingly, four times as great as in the other. And now that I was getting to the root of the matter I wished to take nothing for granted. His equations were high above my head, but I reviewed the position in terms of my own. This is how I set it out: I HAD ALREADY KNOWN THE DIARY NOW TOLD ME That by June 8th he had That his "straphanging" slipped back from forty-five age three weeks later to thirty-five. (on June 29th) was "thirty-three--thirty."