We have seen sense-lines to be a regular feature of Graeco-Latin bilingual manuscripts. They facilitate the scribe's keeping the two texts in step with each other, and the reader's comparing the columns. Another way of achieving this was to write one column evenly, and the other in lines of corresponding content but uneven length. Yet another was to write fewer lines in one column than the other, without exact correspondence between lines. A fourth was to provide an interlinear translation.

TERMINOLOGY

At least three words have been used to describe the manner in which the text of D and similar manuscripts is divided into lines. Scrivener calls them στιχοι (pp.xviif). He is followed by Clark (Acts, pp. xxvii-xli and passim). Kenyon claimed that this was a wrong use of the word, and his opinion has been generally accepted. The evidence is worth reviewing.

There is no doubt about the meaning of στιχος. First it was a line of verse (see Liddell and Scott). It then came to represent, in Kenyon's definition, a 'unit of measurement of fixed length' in prose. This length was shown by Graux to be equivalent to the average Homeric line, calculated as between thirty-four and thirty-eight letters. That a στιχος was the same length as 'the average hexameter' is confirmed by Harris (p. 15) who adds what one would expect, that the measurement is by syllables, not by letters.