The Mayor's Wife, by Anna Katharine Green

Chapter xxi.

The Cipher

Hitherto I had mainly admired Mrs. Packard’s person and the extreme charm of manner which never deserted her, no matter how she felt. Now I found myself compelled to admire the force and quality of her mind, her readiness to meet emergencies and the tact with which she had availed herself of the superstition latent in the Irish temperament. For I had no more faith in the explanation she had seen fit to give these ignorant girls than I had in the apparition itself. Emotion such as she had shown called for a more matter-of-fact basis than the one she had ascribed to it. No unreal and purely superstitious reason would account for the extreme joy and self-abandonment with which she had hailed the possibility of Mr. Steele’s death. The “no” she had given me when I asked if she considered this man her husband’s enemy had been a lying no. To her, for some cause as yet unexplained, the secretary was a dangerous ally to the man she loved; an ally so near and so dangerous that the mere rumor of his death was capable of lifting her from the depths of despondency into a state of abnormal exhilaration and hope. Now why? What reason had she for this belief, and how was it in my power to solve the mystery which I felt to be at the bottom of all the rest?

But one means suggested itself. I was now assured that Mrs. Packard would never take me into her actual confidence, any more than she had taken her husband. What I learned must be in spite of her precautions. The cipher of which I had several specimens might, if properly read, give me the clue I sought. I had a free hour before me. Why not employ it in an endeavor to pick out the meaning of those odd Hebraic characters? I had in a way received her sanction to do so — if I could; and if I should succeed, what shadows might it not clear from the path of the good man whose interests it was my chief duty to consult?

Ciphers have always possessed a fascination for me. This one, from the variety of its symbols, offered a study of unusual interest. Collecting the stray specimens which I had picked up, I sat down in my cozy little room and laid them all out before me, with the following result:

[transcriber's note: the symbols cannot be converted to ASCII so I have
shown them as follows:]

[] is a Square

[-] is sides and bottom of a square,

C is top, bottom and left side of a square,

L is left side and bottom of a square,,

V is two lines forming a V shape

. appearing before a symbol should be inside the symbol

) appearing before a symbol means the mirror image of that symbol

^ appearing before a symbol means the inverted symbol

? is a curve inside the symbol

all other preceding symbols are my best approximation for shapes shown
inside that symbol.

; is used to separate each symbol __________________________

1. []; V; [];.>; V; [-]; <;

2. []; V; [];.>; V; [-]; <; L; ).L; <; )7;.7;

3. []; V; [];.>; V; [-]; <; ).L;.C;[];.L; >;,C; [];.<; ^[-]; ^[-];.<;

4. []; V; [];.>; V; [-]; <; <; L; >; ^V; L; V; []; )L; ^V; [-]; []; V;
).C; ^[-]; >; )C; ),C; V; <; C; ^V; ^[-];.>; [-]; <;

5. *>; []; V; []; *V; []; ~7; )C;.>; ^[o]; )L; ^V; []; Lo; ^V; )C; )7*;
V; )C?; L; )L; 7;.>;.^[-]; )L; >; <;:[-], [-]; Lo;.<;?[-]; )7; [-]; )C;
[];.C; [-]; *7; L;.7; ^V; )o7; *>; C; ^V;.C;.<; [-]; []; 7;.C; )L;:7;
[-]; )*L; C; ^V;.L;.>; ^[%]; C; 7; *L; 7; ):L; )7; ^.V; []; [-];.L;[-]

No. 1: My copy of the characters, as I remember seeing them on the envelope which Mrs. Packard had offered to Mr. Steele and afterward thrown into the fire.

Nos. 2, 3 and 4: The discarded scraps I had taken from the waste-basket in her room.

No. 5: The lengthy communication in another hand, which Mrs. Packard had found pinned on the baby’s cloak, and at my intercession had handed over to me.

A goodly array, if the latter was a specimen of the same cipher as the first, a fact which its general appearance seemed to establish, notwithstanding the few added complexities observable in it, and one which a remembrance of her extreme agitation on opening it would have settled in my mind, even if these complexities had been greater and the differences even more pronounced than they were. Lines entirely unsuggestive of meaning to her might have aroused her wonder and possibly her anger, but not her fear; and the emotion which I chiefly observed in her at that moment had been fear.

So! out of these one hundred and fifty characters, many of them mere repetitions, it remained for me to discover a key whereby their meaning might be rendered intelligible.

To begin, then, what peculiarities were first observable in them?


First: The symbols followed one after the other without breaks, whether the communication was limited to one word or to many.

Second: Nos. 2, 3 and 4 started with the identical characters which made up No. 1.

Third: While certain lines in Nos. 2, 3 and 4 were heavier than others, no such distinction was observable in the characters forming No. 1.

Fourth: This distinction was even more marked in the longer specimen written by another hand, viz.: No. 5.

Fifth: This distinction, which we will call shading, occurred intermittently, sometimes in two consecutive characters, but never in three.

Sixth: This shading was to be seen now on one limb of the character it apparently emphasized and now on another.

Seventh: In the three specimens of the seven similar characters commencing Nos. 2, 3 and 4, the exact part shaded was not always the same as for instance, it was the left arm of the second character in No. 2 which showed the heavy line, while the shading was on the right-hand arm of the corresponding character in No. 3.

Eighth: These variations of emphasis in No. 4 coincided sometimes with those seen in No. 2 and again with those in No. 3.

Ninth: Each one of these specimens, saving the first, ended in a shaded character.

Tenth: While some of the characters were squares or parts of a square, others were in the shape of a Y turned now this way and now that.

Eleventh: These characters were varied by the introduction of dots, and, in some cases, by the insertion of minute sketches of animals, birds, arrows, signs of the zodiac, etc., with here and there one of a humorous, possibly sarcastic, nature.

Twelfth: Dots and dots only were to be found in the specimen emanating from Mrs. Packard’s hand; birds, arrows, skipping boys and hanging men, etc., being confined to No. 5, the product of another brain and hand, at present unknown.

Now what conclusions could I draw from these? I shall give them to you as they came to me that night. Others with wits superior to my own may draw additional and more suggestive ones:

First: Division into words was not considered necessary or was made in some other way than by breaks.

Second: The fact of the shading being omitted from No. 1 meant nothing — that specimen being my own memory of lines, the shading or non-shading of which would hardly have attracted my attention.

Third: The similarity observable in the seven opening characters of the first four specimens being taken as a proof of their standing for the same word or phrase, it was safe to consider this word or phrase as a complete one to which she had tried to fit others, and always to her dissatisfaction, till she had finally rejected all but the simple one with which she had started.

Fourth: No. 1, short as it was, was, therefore, a communication in itself.

Fifth: The shading of a character was in some way essential to its proper understanding, but not the exact place where that shading fell.

Sixth: The dots were necessarily modifications, but not their shape or nature.

Seventh: This shading might indicate the end of a word.

Eighth: If so, the shading of two contiguous characters would show the first one to be a word of one letter. There are but two words in the English language of one letter — a and i — and in the specimens before me but one character, that of [], which shows shading, next to another shaded character.

Ninth: [] was therefore a or i

A decided start.

All this, of course, was simply preliminary.

The real task still lay before me. It was to solve the meaning of those first seven characters, which, if my theory were correct, was a communication in itself, and one of such importance that, once mastered, it would give the key to the whole situation.

[]; V; [];.>; V; [-]; <;

or with the shading (same in bold — transcriber)

[]; V; [];.>; V; [-]; <;

You have all read The Gold Bug, and know something of the method by which a solution is obtained by that simplest of all ciphers, where a fixed character takes the place of each letter in the alphabet.

Let us see if it applies to this one.

There are twenty-six letters in the English alphabet. Are there twenty-six or nearly twenty-six different characters, in the one hundred and one I find inscribed on the various slips spread out before me?

No, there are but fourteen. A check to begin with.

But wait; the dots make a difference. Let us increase the list by assuming that angles or squares thus marked are different letters from those of the same shape in which no dots or sketches occur, and we bring the list up to twenty. That is better.

The dotted or otherwise marked squares or angles are separate characters.

Now, which one of these appears most frequently? The square, which we have already decided must be either a or i. In the one short word or phrase we are at present considering, it occurs twice. Now supposing that this square stands for a, which according to Poe’s theory it should, a coming before s in the frequency in which it occurs in ordinary English sentences, how would the phrase look (still according to Poe) with dashes taking the place of the remaining unknown letters?


A-a —— if the whole is a single word.

A— a —— if the whole is a phrase. That it was a phrase I was convinced, possibly because one clings to so neat a theory as the one which makes the shading, so marked a feature in all the specimens before us, the sign of division into words. Let us take these seven characters as a phrase then and not as a word. What follows?

The dashes following the two a’s stand for letters, each of which should make a word when joined to a. What are these letters? Run over the alphabet and see. The only letters making sense when joined with a are h, m, n, s, t or x. Discarding the first and the last, we have these four words, am, an, as, at. Is it possible to start any intelligible phrase with any two of these arranged in any conceivable way? No. Then [] can not stand for a. Let us see if it does for i. The words of two letters headed by i we find to be if, in, is and it. A more promising collection than the first. One could easily start a phrase with any of these, even with any two of them such as If it, Is in, Is it, It is. [] is then the symbol of i, and some one of the above named combinations forms the beginning of the short phrase ending with a word of three letters symbolized by V [-].<

What word?

If my reasoning is correct up to this point, it should not be hard to determine.

First, one of these three symbols, the V, is a repetition of one of those we have already shown to be s, t, f, or n. Of the remaining two, [-] <, one must be a vowel, that is, it must be either u, e, o, u, or y; i being already determined upon. Now how many [-]‘s and <‘s do we find in the collection before us? Ten or more of the first, and six, or about six, of the latter. Recalling the table made out by Poe — a table I once learned as a necessary part of my schooling as a cipher interpreter — I ran over it thus: e is the one letter most in use in English. Afterward the succession runs thus a, o, i d, h, n, r, etc. There being then ten [-]‘s to six <‘s [-] must be a vowel, and in all probability the vowel e, as no other character in the whole collection, save the plentiful squares, is repeated so often.

I am a patient woman usually, but I was nervous that night, and, perhaps, too deeply interested in the outcome to do myself justice. I could think of no word with a for one of its three letters which would make sense when added on to It is, Is it, I f it, Is in.

Conscious of no mistake, yet always alive to the possibility of one, I dropped the isolated scrap I was working upon and took up the longer and fuller ones, and with them a fresh line of reasoning. If my argument so far had been trustworthy, I should find, in these other specimens, a double [-][-] standing for the double e so frequently found in English. Did I find such? No. Another shock to my theory.

Should I, then, give it up? Not while another means of verification remained. The word the should occur more than once in a collection of words as long as the one before me. If U is really e, I should find it at the end of the supposed thes. Do I so find it? There are several words scattered through the whole, of only three letters. Are any of them terminated by U? Not one. My theory is false, then, and I must begin all over.

Discarding every previous conclusion save this, that the shading of a line designated the termination of a word, I hunted first for the thes. Making a list of the words containing only three letters, I was confronted by the following:

      V   [-]   <

      )L   )C   C

      <    L    >

      ^V   L    V.       <    C   ^V..>.[-]) )L..V   ).C   L.

.<.[-]  )7

      ^V   C    7

      )L.L    >

No two alike. Astonishing! Thirty-two words of English and only one the in the whole? Could it be that the cipher was in a foreign language? The preponderance of i’s so out of proportion to the other vowels had already given me this fear, but the lack of thes seemed positively to indicate it. Yet I must dig deeper before accepting defeat.

Th is a combination of letters which Poe says occurs so often in our language that they can easily be picked out in a cipher of this length. How many times can a conjunction of two similar characters be found in the lines before us..>.[-] occurs three times, which is often enough, perhaps, to establish the fact that they stand for th. Do I find them joined with a third character in the list of possible thes? Yes..> [-] which would seem to fix both the th and the e.

But I have grown wary and must make myself sure. Do I find a word in which this combination of. >.[-] occurs twice, as sometimes happens with the th we are considering? No, but I find two other instances in which like contiguous symbols do appear twice in one word; the.<.[-] in No. 3 and the.V.)C in No. 4 — a discovery the most embarrassing of all, since in both cases the symbols which begin the word are reversed at its end, as witness:.V.)C——— )C.V—.<.[-] —— —.[-].<. For, if.V )C stands for th, and the whole word showed in letters th ——— ht, which to any eye suggests the word thought, what does.<.[-] stand for, concerning which the same conditions are observable?

I could not answer. I had run on a snag.

Rules which applied to one part of the cipher failed in another. Could it be that a key was necessary to its proper solution? I began to think so, and, moreover, that Mrs. Packard had made use of some such help as I watched her puzzling in the window over these symbols. I recalled her movements, the length of time which elapsed before the cry of miserable understanding escaped her lips, the fact that her dress was torn apart at the throat when she came out, and decided that she had not only drawn some paper from her bosom helpful to the elucidation of these symbols, but that this paper was the one which had been the object of her frantic search the night I watched her shadow on the wall.

So convinced was I by these thoughts that any further attempt to solve the cryptogram without such aid as I have mentioned would end by leaving me where I was at present — that is, in the fog — that I allowed the lateness of the hour to influence me; and, putting aside my papers, I went to bed. If I had sat over them another hour, should I have been more fortunate? Make the attempt yourself and see.

Last updated Saturday, March 1, 2014 at 20:37