Encoding text by means of basic ritual magic is one of the first techniques taught to practitioners of necromancy or similar subjects. The stated purpose is to conceal the contents of documents and other assorted writings, but more commonly the ability is used as an introduction to the art. It is simple, requiring a single step, and only includes the most basic of magic energies. Concerning use as an actual cipher, which is to say to protect messages from undesired readings, it is largely unused due to both the presence of superior options and the underwhelming modern need for cryptography. A full examination of encipher, then, only has much merit as a scholarly inquiry - understanding rather than any potential use. Indeed, analysis demonstrates that encipher is on its own not altogether functional.

Fundamentally, encipher operates much the same as any other encryption. It is in essence a process which, being provided with text of arbitrary length(henceforth the plaintext), produces another text of arbitrary length(henceforth the ciphertext). For purposes of demonstration, these may be abbreviated as P and C respectively.

Encipher also requires a key, another text of arbitrary length which should modify the resulting ciphertext and enable accurate production of the plaintext from the ciphertext once produced. As it happens, only the second is in fact a role of the key. It is advised that any ritualists wishing to enjoy this work in the fullest keep handy a writing surface such as the journal the Fates provide. This may serve both to provide independent verification and to enable a more hands-on understanding of the topic.

Proper understanding of a procedure requires proper understanding of the environment in which it is performed. It is then appropriate to begin with an examination of writing. Common as mortals understand it is constructed of words. The nature of a word is a rather more complex than is appropriate for discussion here, so we shall proceed to their components.

Words are written in letters, which are themselves atomic. There are twenty six letters, found in both upper and lower case, summing to fifty two symbols. For the sake of mathematics and general shorthand there are also ten digits, running from zero to nine. This brings us to sixty two symbols. Lastly, in order to separate words and give combinations meaning, there are thirty three symbols which may be described as punctuation.

In no particular order, they are: 
The empty space 
The grave accent, written ` 
The apostrophe, written ' 
The tilde, written ~ 
The exclamation point, written ! 
The at sign, written @ 
The number sign, written # 
The character written $ (see note 1) 
The percent sign ,written % 
The caret, written ^ 
The ampersand, written & 
The asterisk, written * 
The open and closed varieties of parentheses, square brackets, triangle bracket, and braces written ( and ), [ and ], < and >, and { and } respectively
The underscore, written _ 
The dash, written - 
The plus sign, written + 
The equals sign, written = 
The slash and backslash 
The vertical bar, written | 
The colon and semicolon, written : and ; 
The quotation mark, written " 
The comma 
The question mark, written ? 
and the period. 

The total, then, is ninety five unique symbols, one of which(the backslash) has an unfortunate tendency to disappear after being written. The precise cause for this is unclear, the prevailing theory being the existence of particularly selective, ink-eating mites. There is one last set of symbols, those which can be written but cannot be spoken, but they behave quite the same as the primary ninety five with respect to the ritual under consideration and furthermore will never be encountered barring exceptional circumstances, not being generated by enciphering, and are thus unimportant here.

The objects being manipulated understood, the essence of encipher is not terribly complex. Every character is assigned a numeric value ranging from zero to ninety-four, listed below.

Chairman Pejat Shevat's Introduction to Numeric Symbolism contains a more thorough explanation of numeric substitution under Concept 38 and is generally an acceptable guide to the topic it explores. Alas, it proceeds more as an artist's manual on numeraesthetics and ignores entirely the functional importance of primes with regards to the higher magics. Well enough written all the same, of course.

Values being assigned, the remainder of the process is simple addition. To determine the character at a given location, sum the values of the plaintext character and those to its immediate left and right. The beginning of a line behaves as the character i, while the end of one behaves as the empty space. This is done for every character upon the page. The resulting values modulo ninety five(the remainder of the value upon being divided by 95, 100 modulo 95 for instance equaling 5) are converted back to characters, which produces what might be termed the raw ciphertext. 

Lastly, the raw ciphertext is rotated - each character traded for another in the ninety five a certain distance along the set - by a random amount. This random rotation is why enciphering a given text multiple times with the same key can produce apparently distinct results. The astute may have noticed that at no point is the key utilized in the creation of the ciphertext. It is possible that it is used in determining the random rotation, but a random integer added to a fixed integer produces an integer equally random as the first. This is significant because it means manual decryption is straightforward and that the cipher actually uses an incredibly small keyspace: the rotation value, random rotation aside. Furthermore, the key is demonstrably not used in deciphering. Attempting to ritually translate a text with the incorrect key produces a ciphertext which is simply a rotation of the original ciphertext, the value of the rotation being the sum of the characters in the key.

The two keys of `` and a are the simplest significant equivalent keys, both leading to a calculated value of 2. Ritually deciphering a text not encoded with either as the key will result in identical second ciphertexts between the two. However, if either `` or a is the key, then one will dispel the cipher and the other will not. Thus, the key is not in fact utilized in the decryption, but is rather used to trigger the correct dispel.

Concerning manual decryption, the naturally occurring method is simple but troublesome due to the first layer of encryption resulting in precisely ninety five distinct plaintext lines capable of producing identical ciphertext lines. Include the random rotation and the result is that roughly ten thousand distinct plaintext lines may result in an identical rotated ciphertext line. The solution, of course, is to check through them all. One convenience is that once a given line produces which is a likely plaintext, the rotation used there may then also be used for every other line, cutting the possibilities by nearly 99%. Considering the numbers involved, then, some level of labor assignment is likely desirable. Properly tuned crystalline logic units are the easiest solution, but the natural difficulties of acquiring and using them outside of the Collective means that in most cases a number of literate servants will be preferred.

Ultimately, this cantrip is little more than it appears. The simplicity of use, both functionally and in terms of the magical techniques applied, render it well-suited to beginner's practice. However, due to the omnipresence of instant, perfect, uninterceptable-barring a very limited number of exceptions-communication throughout the modern basic, casual encryption is very much unnecessary. Implementation of logic circuit decryption, preferably including successfully interfacing with a dictionary, should be good practice for any students of that particular sort of engineering, but for the rest of the basin, the encipher ritual is unlikely to be of any lasting interest.

[1]. The use and meaning of the symbol $ is, at present, unknown. There are a number of standing theories, the three most popular of which are as follows: first, that it is simply a stylization of the capital S; second, that it is an alternative for enumeration, used in place of the period, sequential numbers, or similar; third, that it is sort of prototype of the treble clef.

Characters in order, starting from value 0 and proceeding to value 94. a-z signifies the lowercase alphabet, A-Z the upper, and 0-9 the digits. _`a-z{|}~ !"#$%&'()*+,-./0-9:;<=>?@A-Z[\]^