Following up on my post "Counterfeit cultural capital" (5/11/2011), David Russinoff sent some additional information about the early history of expressions like "angles comprised of equal right lines" in English translations of Euclid. I reproduce his note in full below, in order to make his efforts available to other interested scholars, while adding a warning to others that this may go a bit deep in the historical-lexicography weeds even for hardened LL readers.
Recall that the OED cites two instances of "comprise" under the meaning "To constitute, make up, compose" (which it characterizes as "rare"), both from 1794. This is consistent with the usage note found at Dictionary.com, according to which this meaning has been around "since the late 18th century". Your search of Google Books turned up two much earlier (and strikingly similar) instances:
"For so tho' a Triangle in the most simple and precise Conception of it be only a figure comprised of three right Lines …" — John Norris's "Essay Towards the Theory of the Ideal or Intelligible World", 1704.
"Seeing then the angles comprised of equal right lines are equal, we have found the angle FDE equal to the angle ABC; …" — Isaac Barrow's translation of Euclid's "Elements", 1714 edition.
Since Barrow's English translation was originally published in 1660, I wondered whether the usage might in fact date back that far, but you also located the first edition on line and found no occurrence of "comprised of". One of your readers then observed that the line in question appeared not in "Elements" proper but in a translation of Euclid's "Data", which is appended to the 1714 edition but not the original.
This afternoon, I killed a couple of hours at my local library and learned that although Barrow translated "Data" into Latin in 1657, it was not included in his 1660 English edition. However, an independent English translation of "Data" was published in 1661 by John Leeke and George Serle, containing the line quoted above (with "comprized" instead of "compriſed"). I took a closer look at the 1714 edition of Barrow's Euclid and found no clue on its title page to the origin of its version of "Data": "Euclide's Elements, The whole Fifteen Books Compendiously Demonſtrated, with Archimedes' Theorems of the Sphere and Cylinder, inveſtigated by the Method of Indiviſibles, by Isaac Barrow, D.D. Late Mafter of Trinity College in Cambridge, to which is added in this Edition, Euclide's Data with Marinus's Preface and a Brief Treatise of Regular Solids. London: Printed and Sold by W. Redmayne in Fewen-ftreet, R. Mount on Tower-hill, and F. and B. Sprint in Little-britain. 1714." In fact, the unattributed 1714 translation of "Data" turns out to be a verbatim copy of the 1661 translation of Leeke and Serle.
Was this 1661 edition an original translation of "Data"? Its preface explicitly lists the inclusion of "Data" among those features "which hath not been done by any that have publiſhed theſe Elements in the Engliſs Tongue …." However, plagiarism has apparently not always been viewed as a serious offense, and one naturally wonders whether the Leeke and Serle usage of "comprised" might have been inherited from a still earlier source. In fact, Thomas Heath noted in the introduction to his 1908 translation of "Elements", "According to [Robert] Potts, [Leeke and Serle 1661] was a second edition of Billingsley's  translation." This is refuted, however, by R.C. Archibald (American Mathematical Monthly, August 1950) and again by Dana Simpkins (Annals of Science, July 1967), who assures us that in Leeke and Serle 1661, "Euclid's Data [was] issued for the first time in England and in English".
It seems, then, that the first two recorded instances of the controversial usage appeared in 1661 and 1704, from two apparently independent sources, and that the first such instance not referring to a thing "comprised of" right lines did not appear until 1794. I'm not sure what to do with this information.
Publish it in a weblog post, what else?