RED AND PURPLE
Once the sixes and their starting point were recovered, it was possible to determine if the language was English or Japanese and to guess plaintext values. The assumption of plain text was facilitated by stereotypic message beginnings. For example, in the following message the sixes were recovered as E-Q-A-D-R-H, and the correct placement of these sixes was determined as:
Cipher: | BRAXE FQCEV QOOXH ECFDL NHQRV QPPLC ERP |
Plain: | .HE.A .A.E. E...E R..E. .REQ. E.... HA.10 |
The correct plain text was postulated to be "The Japanese government requests that ...." From the additional matched plain and cipher, the analysts could attempt to recover the 20-long alphabet.
On 1 May 1939, the Japanese Foreign Office instituted a special code (Phillips Code) in connection with the use of the B machine. Often "text," when finally reconstructed, appeared more like code or a random assortment of letters, than like plain text. While this at first made the problem harder, once the code groups had been recovered the values provided excellent cribs, especially as message beginnings. The following is a typical stereotypic beginning for a message in Japanese which used the Phillips Code.
Cipher: | FGPXP IXUDB DGECZ LBLNU ZQOQH YNMRQ ARJOP DEILO |
Plain: | XFCGJ WFOVD DNOBB FYXFO CFYLC CFMSG TSJVR KHIFI |
Cipher: | AXPPP LIGDK ZDGRA |
Plain: | CGURV FELBK WTLSI |
XFC | = | Number | CCF | = | paragraph |
GJW | = | 15 | MS | = | 3 |
FOV | = | open parentheses | GTS | = | month |
DD | = | 2 | JVRK | = | 16 |
NO | = | of | FIC | = | Begin kana spelling |
BB | = | 1 | FEL | = | End kana spelling |
FYX | = | closed parentheses | GURV | = | Grew |
FOC | = | secret | BKW | = | United States |
FYL | = | additional | TLSI | = | Ambassador |
classification |
- 10 Ibid., p. 2.