As to the third section, there is so little difference in frequency between these letters and some others not included in the high-frequency class, that any distinction found would not be convincing.
The individual cryptogram, however, has happened to contain the sequences VXXU, SRRV, SZZD, DNNV. Application of pointer No. 6 confirms our previous selection of D, V, S, as vowels, and suggests that the letter U might also represent a vowel. Since the frequency of this letter is only 2, we cannot feel so confident in drawing conclusions about it; however, a glance at the contact data shows that it has touched four different letters, which is 100% variety, that one of these four letters is an accepted consonant, and that none of the other three, so far, is an accepted vowel (pointers 3 and 8). The chances are that this letter U, with its low frequency, represents y in some such formation as ally, ully, etty, etc.
With four vowels tentatively isolated, we are now in a position to apply pointer No. 7, and this we may do by returning to the cryptogram and marking for attention each appearance of each supposed vowel. This is usually done by circling each one with a pencil mark. In Fig. 66, a small letter “v” has served the same purpose, and a few serial numbers have been added for convenience of reference. Now let us examine Fig. 66.
At (a), watching the small “v’s,” we find a fairly uniform distribution of vowels except for three long segments beginning, respectively, at the 20th, 27th, and 37th letters. For convenience, these have been copied out at (b). The two of these which are longer, and therefore most likely to contain at least one of the missing vowels, are both found to have included Z and J. Of these two letters, Z is one which was previously discarded (from the central section of the high-frequency group) during our preliminary investigation. Examining it again, to make sure, we find it now as a double between two supposed vowels, and having two additional contacts with supposed vowels (pointers 6 and 8). But J, we find, has never contacted any supposed vowel; it reverses with a supposed consonant, and shows as much variety as could be expected of a letter appearing only three times.
The acceptance of J provides five of the vowels, with frequencies of 10, 9, 6, 2,
| Figure 66(a) v ? v v v v v v v v ? 25 F D R J N U H V X X U R D M D S K V S O P J R K Z v ? v v v v v 50 D Y F Z J X G S R R V T Q Y R W D A R W D F V R K v v v v v v v v v v v 75 V D R K V T D F S Z Z D Y F R D N N V O V T S X S v 80 A W V Z R (b) 20-25 27-32 37-41 O P J R X Z Y F Z J X G T Q Y R W ? t ? t (c) F D R J N U H V X X U R D M D S K V S O P J R K Z e t i y o y t e e a h o a i t h D Y F Z J X G S R R V T Q Y R W D A R W D F V R K e i a t t o t e t e o t h V D R K V T D F S Z Z D Y F R D N N V O V T S X S o e t h o e a e t e o o a a A W V Z R o t (d) Preliminary assumptions: y t e . e a h o a . t h o e t h o . CORRECTIONS: y t O . O a h E a . t h E O t h E . ...to go ahead... ...the other... (e) F D R J N U H V X X U R D M D S K V S O P J R K Z o t i y e y t o o a h e a i t h N F G D W Notify *e**y to go ahead with....... |
and 3 — a total of 30 out of an expected 32. In a longer cryptogram, we should probably look for the sixth vowel among those letters having approximately the correct frequency for making up the expected 40%. As to the present case, we should have no trouble selecting it from the five-letter segment at (b); but this would cause us to spot also the short word in which it is used, and our immediate concern is that of spotting vowels only through their known characteristics as vowels. We will assume, then, that the last vowel cannot be found.
The next step demands that we assign to the most frequent of the supposed vowels the value e, which happens to be a wrong assumption. Concerning this, it may be well to repeat here something which has already been said: In dealing with the simplest of cryptograms, there is often a short detour into trial and error. Also, the average decryptor, accustomed to the work, and fully aware of what he may expect from only 80 letters of text, will usually pause at this point and make some further observations before filling in any of his substitutions. However, there is value even in the making of wrong substitutions; the actual placing of supposed plaintext values in their supposed positions puts the plaintext possibilities before us in visual form, causing us to note easily those very points for which the experienced decryptor examines in advance.
| Figure 67 1. F D R J N U H V X X U R D M D S K V S O P J R K Z D Y F Z J X G S R R V T Q Y R W D A R W D F V R K V D R K V T D F S Z Z D Y F R D N N V O V T S X S A W V Z R. (80 letters). 2. H V X X U T V W D T R A Z D Y F Z J X T V S O U R D S Z R N S E D T S Q X U L K S E V O T D W W V O D R K V T G S R R V T Q Y R A Y M M V A R P V A K D Y X O H V V W K S G G V T J F M S R R K J A R K T D Y M K T J Z K S T O A L R K J F H R K V U G S U M T S F R R K V A Y Q A J O U R K S R U D Y A W D H V D N. (155 letters. - Total for both cryptograms: 235). |
At (c), then, we have made our substitutions. We have assumed that the most frequent vowel, D, is representing e. Having noted the v-v digrams DS, VD, VS, we have selected S, rather than V, as the substitute for a, preferring a digram oa (DS) to a digram ao (VD). This leaves the vowel of second frequency, V, to represent o. This will cause the third of the v-v digrams (VS) to represent oe, not frequent, but better than the digram ao previously mentioned. J, then, probably represents i, and U may represent y.