"During the close of the war," Crossle began at their initial meeting, "our secret wireless stations along the northern sector of the coast were interpreting a series of messages which, for a time, completely baffled every effort at decipherment. The signals always occurred at the same time - three o’clock in the morning - and were transmitted at such terrific speed that they came to us as an utterly unintelligible jargon. Direction finders indicated they were coming from somewhere on the ocean, but the most vigilant patrolling of the Navy failed to disclose the presence of any suspicious vessel."
"Each night we recorded the signals on a phonograph record which was played over and over again in an effort to catch some inkling of the nature of the code. But we had absolutely no results. Then, one day the machine happened to run down and, to our astonishment, the jargon resolved itself into perfectly intelligible sequences of letters. I am going to give you the very last message we received and let you try your skill," Crossle concluded. "Here it is."
SNSZK KXCHR ZAKDC KNMFH STCDE NQS XK ZSHT CDRDU DMSXS GQDD
Russell looked for several minutes at the array of five-letter groups. "How shall I go about it?" he asked, doubtfully.
"There are two types of ciphers," Crossle explained, "substitution and transposition. A transposition cipher is one in which the actual letters of the message have been transposed by a prearranged key. A substitution cipher is one in which the letters of the message have been changed to other letters or symbols. Later, you will quickly discover the difference merely by inspection, but at first you should make a frequency table of each problem in order to determine whether you are dealing with a transposition or a substitution cipher. You make a frequency table by writing the alphabet A to Z and placing a mark after each letter in the alphabet for every occurrence of each letter in the message."
After a few minutes of work, Russell handed over the following frequency table of the cipher Crossle had given him.
A-1, C-4, D-7, E-1, F-1, G-1, H-3, K-5, M-2, N-3, Q-2, R-2, S-8, T-2, U-1, X-3, Z-3
"All right," Crossle said. "Note that K occurs five times, X and Z each three times. From detailed analysis, we find that the letters JKQXZ normally occur with the least frequency in any modern language (less than two percent of the total) while those of the group ETAON occur most frequently. Therefore, the comparatively high frequency of K, X, and Z in this message makes it certain we are dealing with a substitution cipher. To state it another way, these letters, since they occur so frequently, must represent other letters."
"That’s clear enough," Russell admitted. "But now that I know it’s a substitution cipher, what’s the next step?"
"We always assume that any cipher is a simple one until proved otherwise. The simplest known cipher is the so-called Caesar alphabets, in which the letters of the message are represented by other letters, one, two or three spaces forward or backward in the alphabet. That is, A may equal B, B equals C, and so on. To decipher such a message, one merely takes the first letter of the message, in this case S, and continues the alphabet in vertical columns, S,T,U,V, etc. If you will follow this procedure with each letter, the message will be found in one of the horizontal lines."