John Roberts focuses again (Letters, aj 81.12.97) on the only scheme of Goldfinger's of which he apparently has experience - Albemarle Street - and ignores the evidence not only from many other schemes but also from Goldfinger's frequently published words that from the 1920s he designed according to the three rectangles 1:square root of2, 2:3, and Golden Section. An examination of his buildings shows that this is the case, which is fact, not hagiography. The alternative dimensions for Albemarle Street which Roberts gives are at variance with those on the working drawings surviving in the riba Drawings Collection, but even if they were correct the difference is marginal. In view of the exigencies of building and building materials (eg brick dimension), a departure of 1 to 2 per cent from the pure mathematical proportions does not invalidate them, and is well within the latitude which, for example, Rudolf Wittkower allows himself in analysing Palladio's proportions in Architectural Principles in the Age of Humanism.
Incidentally, I should point out that the two diagrammatic errors which I corrected in a letter in the same aj arose in the transcription by others of my drawings on to the aj computer screens.