Natural History

for the character of those times, who here and there had not shrunk from conjectures or even minute additions for the sake of filling out or polishing the discourse. In general, it is appropriate to remember that all the excerptors did not so much wish to transcribe Pliny in the manner of scribes as to make the resources of Plinian erudition conveniently and usefully accessible to readers. From an exemplar of the same stock, brought from Britain into France, the parts once missing—III—IV 5 and VI 88—148, 153—220—were supplied in the Parisian codex Lat. 6795 by a second hand of the XIIth century (E²); and the greatest part of the corrections (E³) that are seen in that codex reveals the same. Likewise, from similar exemplars, other corrections were drawn in the codices FRD, concerning which cf. what shall be said below.

These aids, when rightly weighed, hold the primary place in critical usage, and especially in doubtful passages, unless they can be convicted of error or interpolation by certain arguments, they are judged to be of the highest authority by right. Furthermore, those things that are intact and good in them have been propagated to a great extent even through certain codices of the succeeding centuries—although occasionally obscured by an increasing number of faults—down to that age when editions began to be produced by type, as appears from the readings of the older editions.

II. OF THE RECENT ORDER

The first stock that can be recognized is a lost codex, whose books II—V were thrown into confusion by the transposition of certain quaternions, leading to a strange disorder of the proper sequence—which transposition Detlefsen described and illustrated most accurately in Mus. Rhen. XV p. 270 sqq. 368 sqq. This passed—without the scribes at first being aware of it—into several exemplars derived from it, and Detlefsen concluded from probable indications in Herm. XXXII p. 326 that the same archetype was also once in Britain, even though the Plinian codices of Dicuil, the monk of Ireland (who wrote a booklet on the measure of the earth in the year 825), were too dispersed. From the exemplars of this origin brought into France arose two families of codices, separated by their own lacunae and readings and the variety of other matters, with the agreement between individuals often varying so that it is difficult to define the reason and degree of kinship with certainty.

1. Of the former family are

D: The Vatican Latin codex 3861, a parchment codex of the XIth century, once complete and joined together with the Leiden Vossian codex fol. no. LXI F, which contains XX 186 — XXXVI 97, now torn from it and encompassing only half of the Natural History up to XIX 154.