Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.


  1. Advanced Patent Search
Publication numberUS1318366 A
Publication typeGrant
Publication dateOct 14, 1919
Filing dateOct 14, 1916
Publication numberUS 1318366 A, US 1318366A, US-A-1318366, US1318366 A, US1318366A
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
Table for coding
US 1318366 A
Abstract  available in
Previous page
Next page
Claims  available in
Description  (OCR text may contain errors)




Inventor COSMO FARQUHAR C. FAROUHAR. TABLE Fon comun, otcomns, AND cHEcKING communications.


Patented Oct. 14, 1919.


@Nuol N3 mno mm3 Sno Onno @So @So Sno 23 Se .50 33 NSO Sno QSO momo @omo omo @o3 non@ ono mono hono nono Nono Homo Oono omo @mno mmo @mno mono .vno nno Nono Hno ono Onno @uno Nuno @smo @mno Sno anno Nuno Hno Onno mono mono sono @ono mono omo nono Nono H@no oOQO @mno w25 me 23 n..25 mno $8 N30 H30 93o 2.3 Sno 2.3 25o $3 33 @$0 En@ oi @3o w26 S w3c 33 15o mno m26 Sno @nwo ammo www@ Eno @mno ...Se mno ...N3 N8 Sno om@ @Se En@ Se @Se ...Se Se 23 NS@ S3 OSO @one mono one @one non@ one non@ Non@ Sno 88 @$0 @amo me 3N@ ...amo amo 3m@ Namo 3N@ SNO ammo @mno b@Ho.



SS @0.8 moo o8 m2o No3 S8 02o @moo w26 Soo 26o ...moo 00 30o moo Soo g8 @moo oo S8 @moo @moo moo woo Wmo@ $8 Omoo @5o @Se Koo @So @Se 00 Se NSO Hboo ooo @@oo @@OO booo @@oo nooo h@@oo n @oo Nooo.

Hooo oooo @moo Omoo bmoo @moo @noo noo @moo Nmoo Hmoo Omoo moo @woo woo @.woo voo .woo @woo Noo Hwoo owoo Omoo @moo @noo @noo @noo hvoo @noo Nnoo Hnoo Omoo ONoo @N00 bNOO @Noo mNoo hwNoo nNOO NNoo HNoo ONOO @Hoo @Hoo Hoo @Hoo @Hoo L.Hoo @Hoo NHoo HHoo OHOo @ooo ooo nooo @ooo nooo @Ooo Nooo Hooo Oooo .w ...Hoo





Patented Oct. 14, 1919.


2.3 2.3 3 2.3 2.3 E3 K3 b3 3 Q3 @w3 3 S3 @w3 3 g3 33 @3 $3 83 @n3 mm3 S3 @m3 3 $3 mm3 $3 $3 cm3 $3 $3 S3 23 E3 $3 owOH NANOH HOH odoH onOH ooH bnOH @QOH mnoH hvnoH nnOH NnoH HnoH OnoH oNoH NoH bNoH omOH NOH #NCH nNOH NNOH HNOH ONOH oHOH HOH bHOH wHoH mHOH .wHoH mHoH NHoH HHOH OHoH @OOH OoH bOoH OooH mooH OOH nooH NOOH HOOH OOOH wooo wooo booo,

Hboo oboo @omo wooo bwoo.

Omoo nooo wooo nooo momo Homo Oomo Omoo Omoo bmoo Omoo mwo .wooo nooo Nmoo Hmoo Omoo mHoo HO bwmo oo @woo oo moo Noo #woo Owo momo wooo bnoo Omoo mno wooo nooo Nnoo Homo Omoo @moo moo boo Omoo Omoo No ammo Nmoo Hoo Omoo @Hoo Hoo bHoo @Hoo nHoo Hoo nHoo NHo ,HHoo OHoO @omo ooo booo wooo wooo ooo ^wooo wooo Homo ooo .nboo Nbmo Hboo oboo @ooo wooo booo wooo nooo wooo wooo wooo Homo ooo owwo mo bmoo @woo @moo moo nooo Nmoo Hmoo Omoo owo do bwoo whwoo mmo o nwmo @voo .Goo oo @moo momo bnoo @nwo nooo :doo nooo momo NHoO .nooo


mbmo bmo nbmo Nbmo Hbno obmo mono momo bono momo wooo omo mono momo Homo Omoo momo.

nmo bono @mno momo mno momo momo Hmno Omoo @woo wmo bwmo odmo @voo doo nwmo mdmo Hwmo owho .doo

nu .Hd ad uw ad Hu no .Hm nu am Eo Hm n@ uw 3 :H EH .3.. no .Ho no no Bo .no un HD m9 n: En H9 n Hb. En nh Eh .doo

3 Pm mp 9m .F E E. o op g n ha Mm sa wm ou .um ..5 um un om .ma nm o Mn so ma oo .wn o .mi uo un so nu Mn wn 0.." ma E E 2m on uw n@ hw ME um ma ow ma ..8 @E @o us mm ne S 3 5.o mH om MH 3 @u oH du n H .Aou .Aou


1,3181366. Patented 001.141.1919.


Fl G. 8

01 O2 05 04 05 O6 00 O1 O2 O5 04 05 O6 az Dm cy d1 ex fk gw OO tw 11j vv wi xu yh zt 51 hj iv j i ku 1h mt ng O1 as bf er de eq fd gp 52 os pf qr re sq td up O2 hc io jb kn la mz nm 55 vc wo xb yn za, am by O3 oy p1 qx rk sw tj uv 54 c1 dx ek fw gj hv ii 04 vi wu xh yt zg af br 55 ju kh 1t. mg na of pr O5 i ce dq ed fp gc ho ib 56 qe rq sd tp uc vo wb O6 jn ka. 1 z mm ny o1 px 5 7 xn ya zz ay b1 cx dk O? qk rw sj tv u i vu wh 58 ew fj gv hi ig jh kt 08 xt, yg z s ar be cq dd 59 1g ma nf or pe qq rd 09 ep fc go hb in ja, kz 60 sp to uo vb wn xa yz 10 1m my n1 ox pk qw rj 61 zm a1 bx ck dw ej fv 11 sv ti uu vh wt xg ys 62 gi hu ih jt kg 1s mf 12 zf ae ,bq cd dp ec fo 65 nr oe pq qd rp sc to 13 gb hn ia. j-z km 1y m1 64 ub vn wa. xz ym zy ax 14 nx ok pw qj rv s i tu 65 bk cw dj ev fi gu hh 15 uh vt wg xs yf zr aq 66 it. jg ks 1f mr' ne oq 16 bd cp dc eo fb gn ha. 67 pd qp rc so tb un va. 17 iz jm ky 11 mx nk ow 68 wz xm yy z1 ak bw cj 18 pj qv ri su th ut vg 69 dv e i fu gh ht ig j s 19 we xf yr z e ad bp-V c c 70 kf 1r me nq od pp qc 20 do eb fn ga. h z im jy '71 ro ab tn ua. vz wm xy 21 k1 1x mk nw oj pv qi 72 y1 zx aw bj cv di eu 22 ru sh tt ug vs wf xr '75 y fh gt, hg is jf kr 1e 23 ye zq ap bc co db en 74 mq nd op pc qo rb an 24 fa, gz hm iy jl kx 1k 75 ta. uz vm wy x1 yx zk 25 uw nj ov pi qu rh 76 aj bv ci. du eh ft gg 26 at tg us vf wr xe 77 ha if jr ke lq md np 27 yq zd ac bo cb ydn 7B oc po qb rn sa. tz um 28 ea fz gm hy i1 jx 79 vy w1 xx yk zw av bi 29 kk 1w mj nv o i pu 8O cu dh et fg gs hf ir 50 qh rt. sg ts uf vr 81 je kq 1d mp nc oo pb 31 we xq yd zp ao bb 82 qn ra s z tm uy v1 wx 32 cn da, e z fm gy h1 B3 xk yw zj ai bu ch dt ix jk kw 1j mv ni 84 eg fs gf hr ie jq kd 34 ou ph qt rg sa tf 85 1p mc no ob pn qa rz 35 ur ve wq xd yp zc 86 sm ty u1 vx wk xw yj 36 ab bn ca, dz em fy 87 zv au bh c1'. dg es ff 3'? g1 hx ik jw kj 1v 88 gr he iq jd kp 1c mo 3B mi nu oh pt. qg rs 89 nb on pa. qz rm sy t1 39 5f tr ue vq wd xp 90 ux vk ww xj yv zi ah 40 yc zo an be. cz dm 91 bt, cg de ef* fr ge hq 41 ey f1 gx hk iw j j 92 i d jp ko l o mb nn oa. 42 kv 1 :'L mu nh ot pp; 9 '5 pz qm ry s1 tx uk vw 45 qs rf er te uq vd 94 wj xv yi zu at bg es 44 wp xc yo zb aa bz 95 df e r fe gq hd ip j c 45 cm dy e1 fx gk hw 9 6 ko 1b mn na, o z pm qy 46 ij jv ki 1u mh nt 9 7 rl sx tk uw vj wv xi 47 0g ps qf rr se tq 9B vu zh ag bs cf dr ee 48 ud vp wc xo yb zn 99 ma nz om py ql rx sk 50 Inventor COSMO FARQUHAR IInnlul.





Specieation of Letters Patent.

Patented oet. 14., 1919.

Application led October 14, 1916. Serial No. 125,609.

T0 all whom it may concern.'

Be it known that I, Cosmo Famenne, a citizen of Great Britain residing at Rutherford, in the county of ergen and State of New Jersey, have invented certain new and useful Improvements in Tables for Coding, Decoding, and Checking Communications, of which the following is a specification.

The invention relates to a cipher-book comprising one or more tables for use in coding, decoding, and checking com-munications, the term book being used herein in its broad significance to apply to a single sheet as well as to a group of sheets, and to the latter whether they are bound or loose.

While the invention may be used in connection with all transcribed communications, it is particularly advantageous in connection with cable codes; and in the further c011- sideration thereof its use in connection with cable codes will be described, for with that understanding of the invention its use in connection with other communications will be readily comprehended. I have selected two wellown systems of codes as a means of illustrating the use of my invention Without, however, intending in any way to be limited thereby; one, in which groups of letters are used to indicate a messa e or a part of a message, and the other in w ich groups of numerals are used for this purpose. In theV latter system it is common practice in code books to have equivalent letter desi-gnations for the numerical message cipher 1n order that, if desired, the numerals Vcan be readily translated into letters. lMy invention' contemplates that'when the numeral system is employed this translation shall bez made so that the resultant code message will'l be transmitted by letters rather than numerals. y

The present practice of the cable com panics is to permit a ten-letter combination' to be sent for a unit word price, provided the' combination constitutes a wordwhich fis v pronounceable by syllable. l One diliculty' with the present system is that the `letters comprising the combination,` whether they` are originally used or are the result of a translation from numerals, haveno natural connection between them te assist lin theproper translation whether pronouneeable by syllable or not. As a result, mistakes commonly occur, `both in the letters themselves and in the order thereof. There is a great desirability therefore, for a means by which ten-letter combinations can be made into self-checking pronounceable words, and such is one of the objects of my invention.

Furthermore, if a cablegram has been erroneously transmitted under the present practice, it is necessary to cable back to the sending station that there has been an error in coding or in transmission and to recable the message correctly. This delay and the expense incident to correcting the errors are very objectionable; and a Vfurther object of my invention is the production of means to aid in correcting an erroneous message, without the necessity of recabling the message corrected.

In order that the invention may be the more readily understood, a brief outline of the common practice of coding and decoding messages is now given. In the lettermethod, a common practice is to make use of a combination of five letters as a code symbol to designate the' message desired to be sont, or a part thereof, and to have each of said veetter groups, to which arbitrary meanings have been assigned, vary from the others by two letters. This latter is referred to as a two-letter variation. When a message is to be transmitted, it is reduced to these live letter oups, and if two consecutive groups can e combined into a ten-letter group which is pronounceable by syllable, this is 'done as above indicated. If not a unit word charge is made for eachfive-letter group. For example, it maybe desired t0 send the followin cablegram, Send power, drawing and specification for Greecesend forms of power for assignment ofpatent 4in Austria. `Thefrst part of this :message in a certain code may have been,A 'ven-the 'arbitrary letter-'combination abbo and thesec-i 0nd part', 'theA letter combination debdm. f The combination of these,r 4abbc'cld-ebttm, is obviously unpronounceable by 'syliainfle'v and must be sent as two words. 'l

supposing, however, hat the last lette-ref the iii-st grougliid, be rreplaced by a vom-ll?,y as, ler exaxnp e i, and the lasst Eetteset'the" second fgroup, on,

accomplished by my invention.

-Goneideriug nowl thenumcrclcode method, assume-that it is' desired to *send the followinfg'mescege: Please epp'lyfor an extension ofi-tme-vnoj model lies come to hend?. The

20-irst lpurt of this message in e certain Vnumera -cdde may 'ha-ve 'been given the erbitmny numeral -enibinletion 9834, and the second part "the l numerall combination 04617. In-lthe particular .code considered,-itA may be -fcund-that the ffrst'cfthese groups of-numerels is equivftlet to `the lctter `.group abe'gm/,mid Athe -seccnd-numerahgroup, tothe letterpm. These-combihe-toform the ccabimtion-'mbgmkoptlmobviously an univnounecable cembinetion, which must be'nsentms `-tWo ffwe'ris. 'Hf hoWeveI'-,\ the-:last letter, m, of the rst groupJ be replaced-by, force the vowclva,-end lthe lastflettln mit, of e Sedond'sgroup by, for exemple, the -'vfowel u, tlfe resultant combination #is egakpdu, epronounceablecombinat-ion, which ca-n besent Ifor the mit word.; price. It is 'also commonupmcteein the numeral cipher codes-to thetlwoletter variation i-n t'the equivalent lle-,ter groups so thatfmy inventiencambeapplied thereto 'as --in the cese-@buie lletterclzrlier cedes.

Assume now the *first of `'the illustrwtive code messege ciphers, bbet'debde, has beenreeeived 'e'sffor example, abdez'ebdu: aswill be hilly explained below lthie-ferrer cimfte-fct web 4#byf'frrc-tmis-'ef checking f syem. -Havimgg discovered fths eqnimentsfonmlfl er thmpossiilegreups of two lettere eaohfend thebookis .divided-into talilesifdr fondly iuse. Infone (ff the tdbles are assembled in groups of two, all of the letters which are used in transmitting the message itselfyin -unth'er table are assem- *bled in two-futter groups um letters which are intended either to be substituted for certain of the'letters of the i'uessege cipher r1-oups or to be added thereto, thc groups lieinggvcn series of numerical equivalents: endin-tllev-thirdtebleure |assembled in twolcttcr groups the letters und numerical equiv nlents of the' table first. referred to but in e different manner. This table is primeriiy *intended-towel in detecting the errosof-a wronglytmin'smitted message, end is econveniente lthough not l necessary supplement'to Athe fllrstmentioned table,

flofaid inthellnderstanding of my invention I have illustrated .in the 'drawings which iforme pent ofthis application 'e'xeuiplesof suidltelsles, but/I do not wish to be understood-to' beflirnted to theprecise tables illustrated, es'ether tables, combinations,-'e.nd numerical equivalents embodying my iuvent-ionfmey be devised by those skilled inthe art. 1nt`l1ese drawings- Figures lend: 2 illustreitee. table in 'which all-.ofthe letters ofthe -elphubetere com bimed .-elfphebet-ioally in all of the possible two-lettercombinations l thereof, each colmbiha'tionbei .given-:an arbitery numerical equivalent dliering from that of any-other combination. V:These figures illustrate the table which is intendediprimerijiy for ruse in connection With'the let-ters of-acsblegram designating ther-message itself;

3,4,"5, -6,mnd 7l illustrate el table in which certain -of theuletters of the alphabet arelaT-L eddnfur columns of two-letter groups, with eseries oferbitrery numerical equivalents .-fassigned to the corresponding grupsef; 'earch cUlu-nm,tl1e numerals being arranged --in eol'lmms in numerical border dcwfn'the mimos Ian these figures the letter bolumns were Irepented to permit of illustnaticmherei-n. ITlistafble is prifmafrily forense? for #substieutng'feertn lletters of the clubic message cipher agrowps, ier ter edditioh fthretq-ito: render'fthbm pronounccafble etlz-scl-cheokihg.

' 8 illustrates @e tafble iifn Ywil-lich I the two-letter groups of Figs. l and 2 (ihaving the- :seme l numerical equivalents as in :farid guilas) -hmeermnged--Iincolumns endwlimcs th the numerical] equivalents indicated pirtiy .ebd-Ve- 'nmld i the mide' of the colin-ms. The list'flnumeafls-afbovewthe columns q-ifldiouhes ithe thousands end bu-n1- dreds values of thefgiidwpsinthe columns beneath, 'and thisv numerels-etitheiside otlfie columns indicate the'ftensendmnits luilslucs oiibhe. rourpsinvthenllnes 'md5' ement` thereto. 'miese es'lluet'afte'e tableiwlnich sfntended primarily to aid in detecting the errors o a wrongly transmitted message.

In the table of Figs. 1 and 2 the twoletter groups are arranged alphabetically according to their first letters, and the second letters of the groups are arranged alphabetically with each .letter of the alphaet; and for convenience in use each series thus formed is separated from the others by spaces. Thus in the first series so spaced ofl' there occurs, down the columns, ac, ab, tw, aal, and so on to and including ce; in the second series of groups there occurs 5a, b5, bc, bd; and so on to the end of the series.

Each two-letter group has been given a numerical equivalent and although there is no single two-letter group which has a numerical equivalent as high as 1000, the equivalents are conveniently expressed in four figures; thus aac has the numerical equivalent 495 expressed as 0495; ab, the numerical equivalent 87 expressed at 0087; hy', the numerical equivalent 1 expressed as 0001; az, the numerical equivalent 0 expressed as 0000.

In the tables of Figs. 3 to 7 inclusive there are four columns of two-letter groups arranged with four roups in each line. In the first column eac group is made up of two consonants; in the second column, of two vowels, (considering y a vowel) in the third column, of a vowel followed by a consonant; and in the fourth column, of a consonant followed by a vowel. The twoletter groups in each of the columns are arranged alphabetically down the columns with the same general arrangement as alreadyr described in connection with the table of Figs. 1 and 2. Thus, considering the irst column, there occur the two-letter groups commencing with Z, the consonants combined therewith following in alphabetical order; for example, lb, Zo, Zal, etc. This is followed by the m groups, and the m groups by the roups, etc. For convenience, each o tlgiese letter columns are designated in some suitable way as by the legends, COL 1, Col 2, 001. 3, 001. 4.

The letter group columns are followed by a series of numerical group columns in which there are seventy-five different equivalents assigned tothe two-letter groups of each line of letters. These numerical equivalents start with 0 and advancefnumerically one by one down the columns until they have reached a sufficiently high figure for the purposes of my method, as shown, up to and including 2699. :For convenience in the use of the table, theequivalents are expressed in four digits however large the equivalent may be, thus Z7), au, yl, and ky each have as one equivalent 0 (expressed as 0000) Z0, ce, m, ku, as one equivalent 1, (expressed as 00 1) pg, tu, or, fe, as one equivalent 16 (expressed as 0016); all of these equivalents being taken from the first column. As illustrating the other equivalents, which each of these groups has from the other fifty-nine columns, il), au, y, Icy, each has the equivalents 30 (expressed as 0036), 72 (expressed as 0072), 468 (expressed. as 0408), 900 (expressed as 0900), 1476, 1080, and 2664; Z0, ac, ym, lau, each has the equivalents 145 (expressed 0145), 433 (expressed as 0433), 757 (expressed as 0757), 1369, 1909 and 2593; pg, iu, or, fe, each has the equivalents 268 (expressed as 0268), 844 (expressed as 0844), and 2572.

In thetable of Fig. 8, in which the twoletter combinations have numerical equivalents of the table of Figs. l and 2, the arrangement is made with a view of working from the numerical equivalent to the letter group. Although the table is in two large columns, it is obvious that the second large column follows on from the 'bottom of the first column, and in the discussion thereof would be considered as one continuous large column.

It has already been explained that none of these groups have equivalents in the thousands and that the equivalents of all of the groups are expressed in four figures.

Wlth this understanding, it will be seen that there are seven columns of `two-letter groups. In fthe first column occur all of the two-letter groups which have no equivalents in hundreds; in the second column, all of the two-letter groups which have equivalents in one-hundreds; in the third, all which have equivalents in the .two-hundreds; in the fourth, all which have equivalents in the three-hundreds; in the fifth, all which have equivalents in the four-hundreds; in the sixth, all which have equivalents in the five-hundreds; and in the seventh, all which have equivalents in the six-hundreds. These values are indicated above the columns respectively as follows: 00, 01, 02, 03, 04, 05, and 06. Thus for example the first two nu.- merals of the four-numeral equivalents of as are 00; of bm, 01; of bm, 02;]Lz, 04; an, 05.

Reading the table across the columns, in the lirst line of two-letter groups occur all of the groups which have n-o numerical equivalents in tens or units; in the second line, all which have no numerical equivalents in tens, but the unit equivalent of which is 1; in the third line, all which have no numerical equivalents in tens but the unit equivalent of which is 2, and so on down the column up to and including 99. Thus, for example, the last two numerals of the four numeral equivalent of .ae is 00; of hy', 01; of 5w, 11; he, 71; en, 99.

If the thousands and hundreds values of. the repeated two-letter groups chosen for 'edi illustration, be combined with the tens and unitsv values thereof, theffollowiug four y*riumeral groups are obtained:

az 0000, bm 0211, hz 0471, ,en 0599.

Referring now to the table of' Figs. 1 and 2, it will be seen that these are the numerical equivalents there given.

To make use ofthe table of Fig. 8 in going from numerical groups to letters, the line of figures is referred to4 for the thousands and hundredsl and the columns of figuiresfor tens and units. Assume' for yillustrati on Ithat is desired to find the two-.letter roup equiv, lent of the numeral 0336. Re crring to the line of figures at 'the top of the columns, it is seen that the letter group will occur in the fourth column: referring to the column of figures at the side of the letter groups the numeral 3G is located. The two-letter group desired will therefore lic found in the fourth column in the line opposite 36 and is om. On reference to the table of Rigs. 1 and 2, it is seen that the equivalent of ox there found is 0336.

For convenience in the use of Ithis table, spaces are made 'between the lines 10 ,and 11, 20 and 21, 30 and 31,-and so on by tens: this is obviously ptional 'g Furthermore, it might be noted t at the seventh column is not as long as the preceding six columns, but stops with the two-lettergrou Us and fthe tens and Aunits values 75,., as t e values of the two-letter gro, s ofV igs. 1 and 2 .are then all accounted? or.

The, method oi coding, decoding, and checkingr cablegrams and the use oi the eiphenbook proceeds along the following mes:

Hauin coded themessage'for transmittal rinto t e ordinaryinessage qipher in ten letter combinations composed of two-groups of `five letters each,whether.frorfnra letter code or from a `numeral code translated into letters,` the, last letter o'f each group fforming the tenletter combination is.-'struok out,

,and the first Ifourletters of. each o'fsaid groups are paired o in ,the order found into two.- etter groups. dThe nl uneral equivalents of each of these two-letter roups are then obtained from the table o'f gs. 1 and 2 or similar tables, and added,

Next,A turn to the table of Fis. 3 to7 inclusive, or asimilar table, and ocate the numeral group .which is found by making this addition. The letters to be supplied for those. struck out` from-the, ten-letter combinations will be .found in. the lineiinwlich this numeralgrou is located and ,mustbe chosen Afrom the, our tiro-"letterv 'oups .in that line. If 'the fourth letters lie two Y groups. orming the ten letter combination are both voweis, the twofletter checking groups should be chosen from column 1; if lthe v'fourth letters of the twoletter groups forming *the ten letter combination are both consonants, the' two-letter checking` groups should be chosenv from column 2; if the'fourth letter ot the first group is a consonant and of the second group a vowel, the two-letter checking group should loe chosen from column .3; if the fourth letter of the first group is avowel' and of the second group a consonant,y the two-letter checking group should be chosen from column 4.

H aving chosen th e two letter checking group,

the first ot the two letters thereof is substituted for the letter which Was struck from the end of the first messageoipher and the This leaves-abbo rled, or, divided in accordance with the,above,ab be de bd. The numerical equivalents for; these two combinar tions Vfound in the table of Figs. 1 and 2, added, are as follows:

o?) with the numerical equivalent 0.087

l A(le C (C C i d a cc L u 0067 `Making a total of 0880 Turning now to the table of Figs. 3 to 7 inclusive, this total, 0880, is found about halfiway ldov'vn the tenth numeral column of Fig. 4: and in 4the two-lettergroup line commencingI pg. 'Thelourth letter of each of thetwo original message; cipher groupsare both consonants .so that the two lietter checking .group to he made use .of insupplying the struck-out' letters is iu, and4 t' e messa e her to be sent would therefore be abtcz'lidu.

When a message cipher has been received,A

it` may be decodedfand checked in the reverse ofthe order of coding and transmittal. Toso-e if 'it is correctly transmitted the fifth and tenth letters of the combination vare joined in the `order of receipt in themessage cipher to lforni what is supposed to be the original twoeltter checkin group The remainin letters are roupe in twos, 4in the orderA o receipt inK t e message cipher and their numerical equivalents from the table of'Figs. 1 aud2, er a similar ,table and added. The sum is located in the tables 'of Figs. 3 to'? inclus-ive, or a similar table, and

the two-letter checking groups, of which the, sum is the equivalent, referred to. If the checking grou made up from the message cipher receive( as just stated, is there found, the checking sho-ws that the cablegram was correctly transmitted, and the further decoding is then proceeded with in the ordinary manner except that the fifth letter of each of the live-letter groups is not found in the message cipher: but as in the systems of codes assumed for illustration, there is .always a two letter variation: this does not affect the decoding, which can readily be done from the first four letters of each fiveletter roup.

Obviously in the systems of codes where the four-letter groups are employed, either originally or as numerical equivalents, the two-letter variation is not necessary and the same general procedure above outlined would be employed, except that instead of replacing letters, the two-letter check group would simply be added, the first letter thereof at the end of the first four-letter group, aud the second at the end of the second fourletter group.

Assuming, however, that in the cipher message desired to be sent there is an uneven number of groups so that there is an odd group of five or four letters, depending on the code system employed. The invention is applicable to that condition as can readJ ily be understood; for instead of there being four two-letter groups for addition to ascertain the checking total, there are butV two two-letter groups, and the last letter of an odd five-letter group is replaced by the checking two-letter group, or the last named group is added to an odd fourfletter group 1n accordance with the cipher system employed.

Thus, as an illustration, assume that the entire message desired to be sent is expressed by the message cipher ribbed. Proceeding in the same manner as before, the numerical equivalents of the first four letters of the group are found in the table of Figs. 1 and 2 as follows:

ab with the numerical equivalent 0087 b c (C G t i Making a total of 0461 Referring to the table of 3 to 7 inclusive, it 1s seen from Fig. 3 that 0461 is in the line commencing with sk. Thev Y j propernumerical equivalents, if t checking' equivalents of the cipher word correctly transmitted totaling to the Same sum as the numerical equivalents of the twoletter checking group, there will be a discre ancy. Obviously the error may be eit er in the message cipher proper or in the checking group and the problem is thus twofold: to discover, first, where the error is, and second what the error is.

The method of procedure can probably best be understood by first giving an illustration. Assume the same exam le as Was referred to in the preamble hereo z'. e., that the message cipher abbcz'debdu had been received as abdcdebdu, that is to say, that the third letter b had been received as CL The total of the message cipher proper equivalents ascertained from the table of Figs. 1 and 2 as before, is obtained thusab with the numeri cal equivalent 0087 da i( 4 C de c: u t c: 0352 bd c: u u c: 0067 Making a total of 0773 Vgroup with this numerical equivalent is either plc y os or fa. thereforewrong.

Assume that the error was in the meage cipher proper and that the checking group u iscorrect. The closest u value to 0773 is :0772. The total of the message cipher at is` the particular value of u intended, is 1 'in excess ,ofmthe correct value. The numerical equivalent of one of the two-letter groups of the message cipher proper is probably, under` that assumed value of z'u,`0001 too high, andthis 0001 should be subtracted from each of the groups Asuccessively (or until ,the errorlis found) to ascertain if the resultant two-letter groupfhavin A the rectiyfied numerical equivalent islik yto be the one in which th'e error occured. Thus,`the correct numerical equivalent ,of the .first two-.letter gronpmay havebeen 0086.A ARelferring to theftable of Fig. A.8 to ascertain The cipher received is iiifiy me ege reeejivee teefeeefleen ber ifi-ime il? wien the errer-'corrected on i this assumptidn, the decoded message iee likely one to be received it is fair toessume thatthe error has heen discovered and corrected. This is found to be the case in the particular exemple cited The general rule,` therefore, when the checkin shows that; e message cipher is incorrect, is to filet endeavor to decode the message cipher proper' vto see if the' decoded message is one likely tobe received. If the erroris in 'the message cipher :pt-oper; this would be shown et once hy' the decoding: either" the` decoding 'will be found to be impossible, or the decoded messa. e Will neketnosense, or will be one not'li ely to be"re`ceived. When any of these are true, it is` fir to assume that the check letter group'is coivrect. Then take various numericl equivalents of the check letter group from the-table ofV Figs. 3 to 7 inclusive or u isimilar table and add or subtract, es the case may he, the diierence between these individuel numerical Iequivztlc'ifits and the sum' oftho equivalents ofthe twodetter groups in the message cipher pI-oper, to the numerical equivalents'ofeech rof the twolettei' groups of 'the message'cipher 'proper (or i until* the error 'has been discovered): eecertin from-tlietble of Fig; 8 orasimilar vtdble'lthe two-letter vGfroup 'of `which this sunror di'erence' is' te equivalent; and continue with theprocese until a twodetter grou' lilas been "diecoveredg -for whiieh'lthe twr-I 'ett'ef guet-1p ofthe erroneous message" iseotkblythefsebsteete. Rewrieeend deco e the' meesa'gefoP-the assumption.' If the rsultent idelsoded" message is'one likely to 'be reeived it Vis? farto assume tht the errer hes' been discovered. There isawpossibility that the process will result in d1scoverng'mol-e than one twoelether groupwhich miglithelveibeeu seiit'in Jche erroneous `message cipher received. l In such :.1 case the sense' of' they meesegwendother elemente later te be* diecussed will' enable' ythe receiver te determine -wihihof theitwollebter Igroups is the correct one.

le @e the' @eher heed, the Ifirst decoding o fthe message results in one which Waf=115 l' ely to be received, it is fair to assume thtthe QHQFWQS in the checking group, 1n such a cese, take the sum of the numerical are adopted and made use of, the parties employing them usually do so in particular matters or lines of business and thatl when code messages are thus employed the receiver of the message can usually tell Whether the message is one which the sender would likely send: furthermore, codes are generally worked out so that an error 4would result in a cipher impossible of decoding or a decoded message which would not make sense. Consequently the sense of the message is a factor upon which considerable dependence can be placed. In making use of the table of Fig. 8 or a simlar table and in endeavor-ing to ascertain what are the correct letters, in addition to the similarity of the letters, the pronounceability of the ten-letter group with the assumed corrected letters therein, and the question whether the second and fourth two-letter groups end with a vowel or a consonant and consequently whether the checking group would be chosen from the first, second, third or fourth columns of the table of Figs. 3 to 7 inclusive, are matters of considerable assistance, and when taken in connection with the sense of the message result in the figurincr out of the correct message.

ecause of the fact that the two-letter groups and the numerical equivalents in the tables of Figs. l and 2, and 8 are identical,

either of these tables (probably the table of Fig. 8) might be dispensed with, without departing from the spirit of my invention, but dispensing with either one of these tables is not advised because of the resultant great inconvenience.

The foregoing detailed description has been given for clearness of understanding, and no undue limitation should be deduced therefrom, but the appended claims should be construed as broadly as permissible in view of the prior art.

What I claim as new and desire to secure by Letters Patent off the United States is:

l. As a new and improved article of manufacture, a cipher-book comprisin a suitable number of pages containlng ta les for use in coding, decoding and checking communications, one of said tables oomprising in suitable arrangement groups of letters with arbitrary numerical equivalents for use in the coding and decodin of the communications, and the other of said tables comprising in suitable arrangement groups of letters With numerical equivalents for use in connection with the letters of the message cipher for the purpose of checking the coded communications, substantially as described.

2. As a new and improved article of manufacture, a cipher-book comprisin a suitable number of pages containin ta les for use in coding, decoding and ciecking communications, one of said tables comprising in suital'ile 'arrangement groups of letters, each group having arbitrary numerical equivaients for use in the coding and decoding of the communications, and the other ot said tables comprising in suitable arrangement groups of letters, each group having a series of numerical equivalents for use in connection with the letters of the cipher message for the purpose of checking the coded communications, to the end that pronounceable self-checking cipher communications may be produced, substantially :is described.

3. As a new and improved article of manufacture, a cipher-book comprisiiw' a suitable number of ages containing tables for use in coding, Idecoding and cl'iecking communications, one of said tables comprising in suitable arrangement two-letter groups, each group having an arbitrary numerical equivalent for use in the coding and decoding of the connnunications, and the other of said tables comprising in suitable arrangement twodetter groups, cach group having a series of numerical equivalents for use in connection with the letters of tlie cipher message for Jthe purpose of checking the coded comn'iunications to the end that pronounceable selfchecking cipher combinations may be produced substantially as described.

fi. As a new and improved article of manufacture, a ci her-book comprising a suitable number oi pages containing tables 'lor use in coding, decoding and checking communications, one of said tables comprising in alphabetical arrangement two-letter groups, each group having an arbitrary numerical equivalent for use in the coding and decoding of communications, and the other of said tables comprising in alphabetical arrangement two-letter groups, each group having a series of numerical equivalents or use in connection With the letters of the cipher message for the purpose of checking the coded communications, to the end thatpronounceable self-checking cipher communications may be produced, substantially as described.

5. 'As a new and improved article of manufacture, a ci her-book comprising a suitable number oiP pages containin tables for use in coding, decoding and c ecking communications, one of said tables comprising in alphabetical arrangement two-letter groups, each group having an arbitrary numerical equivalent for use in the coding and decodin of the communications, and the other o said tables comprising two-letter groups arranged in four parallel columns, the groups proceeding alphabetically down the columns, the corresponding groups in each column havin a series of numerical equivalents, the sai groups bein for use in connection with the letters of t e cipher xovasagc for the purpose of checking the coded communications, to the endr that pronounccabic soif-checking cipher connuunications may bo produced, substantialiy as dc- :-(-rii cd.

G. As a new and improved article of manufacture, a; cipher-book cornprisiur a .suitable nulnbcr oi pages containjr ta los For use in coding, decoding and ibicck-ing connunnications, one of said' tables comprising; in aiphabetiical arrangement in columns` two-letter groupe, each group having. an arl'iitrary numerical equivalent, the said groups being for usey in the coding and dccoding of communications; one of said tables comprising two-letter groupe4 arranged in four parailbi columns, the groups proceeding alphabetically down the columns, the corresponding groups in ouch column havingl a seriezI of numerical'equivalents, thc said groups being' for use in connection with tho icfi'ors4 of ihr cipher nleesagc for the pnl-- pose of ciu-cking the codcdf connnuniearions; und the othcr of Huid tabics conlpiis ing the tivod'ettcr groups and the ninnerimi c |ui\'aicn1s o' t'hc rct named liable :1r ranged in numerical order; to the end that ln'onounceable Self-checking cipher coinmunications may be, produced and if the same contain errors, that the said errors may bc corrected. substantially as described.

7. As a new and ini-proved article oi' mannfact ire, a cipher-book compr-isili a, suitable nun, er oil? pagos containing. tab 'es for use in coding', decodl'iiigland checking communications, one of said tables comprising in alphabetical arrangement inV columns twoletter groups, cach' group having an arbitlfary numerical c uiwaient, the Said? groulis being for use in 1e coding and dfecoding ol" connuunlczugions; one of'said tables comprising two-leiter roups arranged in four arall'el columns, ic groups proceeding alphabetically down the columns, the corre Spontini@ groups inyeach coiunm having a series o numerical' equivalents, ifhe said roups eing Afor use in' connection with ge letterscf the clf-pher message forthe purpose olf'checlrirw t 1"?. coded' communications; and' theJ ot'iie, Said tables comprising in a phnaIny of columns the (info-letter groups und' the' nlnneljicuil equivalents of Liie irstinamed lfablc arranged in nlnncricai order, tjiw numerica] equivalent of cach group hl ing o qnesscril partly above the colfnlnns: of f groups and partly attihc l@idc ifhereof, to thc vnd that, pronounceaiiic Scifi-checking cipher connnunicatiom may be produced and if the carne contain errors @hat the said errors ma f be corrected, eubsijantiallylas described.

i ign'ed lat New Yorii, in thev county of New" ori; andSateof New York, this 11th day ofct., 1916.


fig ents eagi, Tay addesqig the Commissioner off Patents, www@ I? e"

Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US4957298 *Mar 22, 1989Sep 18, 1990Jack SilvermanPermutation group games
U.S. Classification283/17, 380/59
Cooperative ClassificationB42D15/0013