Terminus: a copy and imperfect mirror.
In his 1848 Gromatici Veteres (a title which makes me think of Nick Park’s stop-motion animation), Lachmann relied on the famous (and obviously ancient) Codex Acerianus. But he also used another manuscript from the former ducal library at Wolfenbüttel (Cod. Guelf.105 Gudianus Lat.) as a principal source, and relegated the collations of Vatican Pal. lat 1564 to the variant readings at the bottom of the page. Lachmann’s favored manuscript was from the collection of the German philologist Marquard Gude (1635-1689), and was acquired for the Herzog August around 25 October 1709 (I think), by Gottfried Wilhelm Leibniz. Quite simply, Codex Gud.105 is a copy of Pal. lat. 1564 from late in the ninth century, but the correct order of descent was not decisively established until the Swedish scholar Carl Olof Thulin (1871-1921) published his Die Handschriften des Corpus Agrimensorum Romanorum in 1911 (ibid.58-69).
Of course, he used methods pioneered by Lachmann decades before. Although the paths the two codices took through the passage of centuries were quite distinct, we can witness the imperfect mirror formed from an image and its copy. Unfortunately, few current digitization projects provide us with tools to explore these paths of transmission, circulation and reconstruction. On this point, I will not yield.