My Web page about Cryptology
The Louis XIV's codebooks
|
IntroductionIn Catinat's memoirs published in 1819, a double-table code is presented. It corresponds to the printed version of a manuscript present in the correspondence of the Marshal and kept by his family. This code would have been used for correspondence between Louvois and Mr de Bonnais for the year 1676. Until proven otherwise, it is the oldest known double-table code. The codebookHere are the two tables that make up this codebook:
Table to encode, for the month of October 1676=====> A fi 159 277 le 147 288 pi 43 247 to 59 160 A 116 200 292 fo 26 93 li 177 248 po 140 275 tu 13 165 =====> B fu 141 255 lo 20 191 pu 184 256 tout 122 271 B 14 169 251 faire 183 234 lu 48 124 par 7 281 =====> V ba 73 300 =====> G leur 227 pour 53 107 V 32 148 253 be 31 138 G 51 175 252 lui 66 =====> Q va 172 268 bi 241 299 ga 81 225 =====> M Q 21 164 293 ve 25 187 bo 100 250 ge 163 289 M 12 103 157 qua 83 126 vi 133 236 bu 170 239 gi 194 257 ma 86 198 que 211 222 263 vo 144 238 =====> C go 34 245 me 143 259 qui 9 60 vu 49 196 C 50 137 298 gu 134 207 mi 91 296 quo 104 181 vostre 92 206 ca 58 188 =====> H mo 24 131 qu 94 278 vous 139 290 ce 6 89 H 151 217 295 mu 52 223 =====> R =====> X ci 201 260 ha 4 61 monsieur 202 R 33 101 276 X 27 72 258 co 79 130 he 155 232 =====> N ra 90 240 xa 109 214 cu 1 229 hi 98 237 N 3 99 242 re 39 221 xe 142 233 =====> D ho 186 272 na 46 270 ri 117 249 xi 29 115 D 44 101 269 hu 76 128 ne 8 123 ro 57 209 xo 192 284 da 2 110 homme 113 243 ni 215 291 ru 174 261 xu 67 112 de 150 228 280 =====> J no 28 153 =====> S =====> Y di 179 205 J 30 111 266 nu 80 265 S 17 162 168 28 Y 176 do 22 145 ja 56 108 nous 36 244 sa 199 273 =====> Z du 213 286 je 166 285 nostre 63 se 15 129 Z 35 118 216 =====> E ji 5 132 =====> O si 231 297 za 23 106 E 10 64 149 jo 96 210 O 16 82 267 so 47 120 ze 97 189 est 121 274 ju 161 279 ons 54 171 su 218 282 zi 65 254 et 38 84 =====> K ont 180 294 =====> T zo 178 219 =====> F K 87 173 =====> P T 19 75 193 zu 37 152 F 105 220 283 =====> L P 85 136 204 ta 70 230 =====> Nulles fa 18 68 L 41 119 262 pa 11 114 te 88 246 Nulle 40 55 62 7 fe 125 264 la 71 182 pe 77 235 ti 45 190 Table to decode, for the month of October 16761 cu 50 c 100 bo 150 de 202 monsieur 254 zi 2 da 51 g 101 r 151 h 203 nulle 255 fu 3 n 52 mu 103 m 152 zu 204 p 256 pu 4 ha 53 pour 104 quo 153 no 205 di 257 gi 5 ji 54 ons 105 f 154 nulle 206 vostre 258 x 6 ce 55 nulle 106 za 155 he 207 gu 259 me 7 par 56 ja 107 pour 157 m 209 ro 260 ci 8 ne 57 ro 108 ja 159 fi 210 jo 261 ru 9 qui 58 ca 109 xa 160 to 211 que 262 l 10 e 59 to 110 da 161 ju 213 du 263 que 11 pa 60 qui 111 j 162 s 214 xa 264 fe 12 m 61 ha 112 xu 163 ge 215 ni 265 nu 13 tu 62 nulle 113 homme 164 q 216 z 266 j 14 b 63 nostre 114 pa 165 tu 217 h 267 o 15 se 64 e 115 xi 166 je 218 su 268 va 16 o 65 zi 116 a 168 s 219 zo 269 d 17 s 66 lui 117 ri 169 b 220 f 270 na 18 fa 67 xu 118 z 170 bu 221 re 271 tout 19 t 68 fa 119 l 171 ons 222 que 272 ho 20 lo 70 ta 120 so 172 va 223 mu 273 sa 21 q 71 la 121 est 173 k 225 ga 274 est 22 do 72 x 122 tout 174 ru 227 leur 275 po 23 za 73 ba 123 ne 175 g 228 de 276 r 24 mo 74 nulle 124 lu 176 y 229 cu 277 fi 25 ve 75 t 125 fe 177 li 230 ta 278 qu 26 fo 76 hu 126 qua 178 zo 231 si 279 ju 27 x 77 pe 127 nulle 179 di 232 he 280 de 28 no 79 co 128 hu 180 ont 233 xe 281 par 29 xi 80 nu 129 se 181 quo 234 faire 282 su 30 j 81 ga 130 co 182 la 235 pe 283 f 31 be 82 o 131 mo 183 faire 236 vi 284 xo 32 v 83 qua 132 ji 184 pu 237 hi 285 je 33 r 84 et 133 vi 185 nulle 238 vo 286 du 34 go 85 p 134 gu 186 ho 239 bu 287 s 35 z 86 ma 135 nulle 187 ve 240 ra 288 le 36 nous 87 k 136 p 188 ca 241 bi 289 ge 37 zu 88 te 137 c 189 ze 242 n 290 vous 38 et 89 ce 138 be 190 ti 243 homme 291 ni 39 re 90 ra 139 vous 191 lo 244 nous 292 a 40 nulle 91 mi 140 po 192 xo 245 go 293 q 41 l 92 vostre 141 fu 193 t 246 te 294 ont 43 pi 93 fo 142 xe 194 gi 247 pi 295 h 44 d 94 qu 143 me 196 vu 248 li 296 mi 45 ti 95 nulle 144 vo 197 nulle 249 ri 297 si 46 na 96 jo 145 do 198 ma 250 bo 298 c 47 so 97 ze 147 le 199 sa 251 b 299 bi 48 lu 98 hi 148 v 200 a 252 g 300 ba 49 vu 99 n 149 e 201 ci 253 v
Is this codebook the oldest with double tables?The presence of this code in the papers held by Marshal Catinat is very bizarre. It should be noted that it was accompanied by security measures (described in the security chapter) sent to Mr de Bonnais dating from 1677. At the time Catinat was not very senior, he was only a lieutenant in the regiment of guards. I don't think he used secret codes at that time and a priori he has no connection with Mr de Bonnais. Here is my hypothesis: When Catinat reaches the highest office (Captain in 1679 and Marshal in 1680) he is required to communicate in encrypted form. We send him the measurements security to take and he is given an example of a code (and perhaps an example of a cipher letter?) which are copies of the information transmitted to Mr de Bonnais a few years ago. If I am right, other copies of these same documents may be found in other archives. Moreover this accredits the fact that this code would be the first to use two tables. Analysis of the structure of this codeIt is small, it includes less than 300 groups. It is made of two tables, one to encrypt and the other to decrypt. This code is very simple. It mainly contains the letters (a, b, c, …) and the syllables (ba, be, bi, bo, bu, …). We recall the confusion of the letters i,j and u,v and the absence of the w (vv). There are also some very common words in the French language: est, faire, homme, et, lui, lui, monsieur, nous, nostre ("notre" in Old French), par, pour, tout, vostre ("votre" in old French) and vous (in all only 14 words). Finally, there are only two suffixes: ons and ont. There are no proper names. Unlike the previous single-table code of the same year (1676), the cipher groups for letters and syllables, and some very frequent words, are each represented by several cipher groups: The letter « r » : 151, 217, 295 The syllable « da » : 2, 110 The word « faire » : 183, 234 Finally, the codes numbered from 1 to 300 do not all have a meaning. For example groups 42, 69, 78, 102, 146, ...226 (in all 13 groups) do not exist (they have no clear correspondence). My hypothesis is that these absences make it possible to end the code exactly on the codegroup of value 300. References
|