When I had arrived at this result, and had tested its accuracy in working, I felt in a position to experiment with my new knowledge on the old number cipher. First I wrote out my method of reduction as a sort of addendum to the paper which I had prepared for Marjory. Then I made a key to cipher and one to de-cipher.[3] By this time the night was well on and the grey of early morning was beginning to steal in by the edges of the blinds; I was not sleepy, however; I was too much excited to think of sleep, for the solving of the problem seemed almost within my grasp. Excited to a state which almost frightened me by its intensity, I got ready my copy of the number cipher and my newly prepared key. With an effort which took me all my resolution I went on steadily writing its proper letter under each combination without once looking back; for I knew that even should some of the letters be misplaced in the key the chance of recognising the right ones would be largely increased by seeing a considerable number of letters together. [3] See Appendix C. Then I glanced over the whole and found that many of the symbols made up letters. With such a basis to work on, the rest was only labour. A few tentative efforts and I had corrected the key to agreement with some of the combinations in the cipher. I found, however, that only here and there were letters revealed; try how I would, I could not piece out the intervening symbols. At last it occurred to me that there might be in the paper two or more ciphers. On trying to follow out the idea, it became apparent that there were at least a quantity of impeding numbers scattered through the cipher. These might be only put in to baffle pursuit, as I had surmised might be done when I made the cipher; or they might have a more definite purpose. At any rate they hampered my work, so I struck them out as I went along. That I continued till I had exhausted the whole list of numbers in the script. When I looked back over the letters translated from the cipher thus depleted, I found to my inexpressible joy that the sequence and sense were almost complete. The translation read as follows: "To read the history of the Trust use cipher of Fr. Bacon. The senses and the figures are less worthy than the Trinity B. de E." One step more and my work was done. I set the discarded numbers in sequence on another sheet of paper, and found to my intense satisfaction they formed an inner record readable by the same key. The "encloased" words, to use Bacon's phrase, were: "Treasure Cave cliff one and half degree Northe of East from outer rock." Then and then only did I feel tired. The sun was well up but I tumbled into bed and was asleep in a moment.