"Time went on, and the solution was as far off as ever; at least apparently. Little Hugh and I had come to Irvington for the winter; it was close to Christmas, and I had the blues terribly. Just to think of Christmas and that abyss lying between us! For I knew that you would not come unless I called, and I could not send for you quite yet. Suppose that the discovery of the secret should be close at hand; I might need Chalmers to help out on some difficult scientific point.
"It is always the little things that show the way out. Hilda's weekly letter had come, and I was reading it eagerly hoping to find some mention of you. Now Hilda, poor dear! is an awful speller; she never could learn to visualize words. As I read along I came on a word which looked odd; then I saw that she had committed the careless stenographer's error of spelling 'forty' with an u, thus: 'fourty.' Of course, the pronunciation is the same in either case—and then it was that I got my big idea. Was it possible that the phonetic sounds in my series of numbers might fit words of entirely different meaning than their ordinary equivalents in letters? Let me try.
"1-4-2-4-8. Why, yes, 1 is 'one' and also 'won'; 4 is 'four' and also 'for'; 2 is 'two' and also 'too'—quick! let me get them all down. And here was the result: Won—for—too—for—ate. You see that, in every instance, the phonetic sound of the number can be represented exactly by a word of entirely different meaning. But this peculiar quality in the series, 1-4-2-4-8, would not be apparent at a casual glance, and the figures could even be written down for future reference, or sent to a distant correspondent, without any probability of that inner significance becoming revealed. Very clever of Fielding Thaneford—that is if my deductions were really correct!
"The first step was to set down the new key-sentence with the cypher writing underneath. Here it is; this time using fifteen letters."
| W | O | N | F | O | R | T | O | O | F | O | R | A | T | E |
| Q | W | O | T | T | U | I | J | X | I | S | V | A | Z | P |
"Applying the decoding rule I got the following in my first six tries:"
| T | H | A | N | E | C |
"You can imagine how excited I was. If my theory were correct the next four letters should be OURT, completing the word 'Thane Court,' Eureka! it is coming! It is coming! I got both the O and the U.
"From the height of exultation to the depths of despair. For instead of R in the ninth place, I had to set down an I; and then, in succession: CDD-FKL. Perfectly impossible! Look at it: THANECOUICDD-FKL, etc.
"And yet the cypher had certainly started to uncode; what could have thrown me off the track? For I had succeeded in getting 'Thanecou,' and that unusual combination was significant in the highest degree. What word could it be but 'Thane Court,' the ancestral home of the Thanefords? Why the chances were a million to one against my reaching such a series for—for——"