^{page 178 note 1} Italics throughout are mine unless I specify otherwise.

^{page 178 note 2} Mr Housman says I ‘essay to base a conjecture on an unsupported reading of L at III iv 22.’ What I essay to base it on is that reading taken together with the serious discrepancy media and sacra between the other MSS. He again perverts my argument on this passage in discussing II xxiv 46. To make unlikes like by confusing the circumstances seems to be out of many Mr Housman's favourite paralogism.

^{page 179 note 1} Mr. Housman uses the indications of F at xxii 22 and elsewhere in a precisely similar way J.P. xxi 140.

^{page 179 note 2} In his note on peraeque III xxviii 9 (p. 27 n.) Mr Housman, angry at being told he has not attended enough to manuscript abbreviations, surpasses himself. L (L writes e for ae always) having peque, D (and seemingly V formerly) pareque, F paremque and N per aequae and I having said the original of all was ‘neither “pereque” nor “pareque” but “peque”,’ Mr Housman puts forth as ‘probable’ the peculiar hypothesis‘that the readings of DVF are due to peque, which their common parent O had misinterpreted as pareque.’ Why? To deprive L of the honour of having preserved the common tradition more faithfully than DV and F. As to N's per aequae, I had derived that from veque because, as the variation of ae and e is unimportant and is disregarded here by Mr Housman himself, J.P. xxi 131 ‘per aeque, that is peraeque,’ there is no solid ground for separating N from the other MSS with which it agrees far oftener than it differs. But Mr Housman says ‘N derives scores of readings from an older source than O, and there is not a hint that this source had geque rather than peraeque.’ So, having provided one source older than O with -peque against L, he provides another with peraeque against me. Place aux ombres! After contradicting my statement that N's reading was no better than DV's or F's (FI added, speaking to one point at a time) by saying that ‘truth is more obscured by wrongly expanding the abbreviation’— let him add, to himself—‘ than by misdividing the word’ as if that were the test to apply to a scribe, he ends with this outburst:‘On p. 23 an equally baseless charge of neglecting abbreviations is brought against Mr Leo and supported only by flat contradiction.’ The wondering reader may be told that Mr Housman has discussed the place in question (II xxxii 8) in J.P. xxi 174 without referring to the abbreviation (tⁱ for tibi), a compendium also ignored ib. p. 129, on IV iii 51 and ib. p. 141, on IV v 85: compare my pamphlet p. 23 and n. 1.

^{page 180 note 1} ‘For no discoverable reason unless it is the hope of boasting “liquidisimmisifontibusapros”’ (p. 22).

^{page 180 note 2} Of the contumelious reference here to my discussion of III xiv 19 (p. 184 below) I intend to make Mr Housman ashamed.

^{page 181 note 1} The matter I add in square brackets is taken almost verbatim from Mr Housman, p. 21.

^{page 181 note 2} In this connexion I had written that ‘Mr Housman's advocacy of the value of F's isolated witness involves him in a curious inconsistency. He follows Baehrens in maintaining that “A is the most faithful representative of its family ”… It certainly then “happens curiously” that in the poems in which we have both A and F, A should give of itself but one true reading…and F three or four,’ which Mr Housman thinks after all may be an accident. His defence, with Dick, Tom and Harry to the rescue, is ‘wherever A gives a true reading that reading is also given by F or by N’ (hisitalics)‘or by both.’ Those acquainted with the laws of probability will still think it curious that the faithful A, uplike the less faithful F, should never be right by itself. But the imputation of ‘inconsistency ’I will withdraw, as I can find no evidence that Mr Housman has any such acquaintance.

^{page 181 note 3} Mr Housman's reference to these instances gives cnrious indication of the bias and confusion in his mind. ‘In four of them F is now confirmed by L, so Dr Postgate must confess that in those four I was right,’p. 21. Why, when did I say otherwise? I never criticized his views on F till the pamphlet he has attacked; from it alone does he derive these confirmations and even the remark is ‘plagiarised’ from me, p. 29. He should have written ‘In four of them F is now confirmed by L, as I have learnt from Dr Postgate.’

^{page 181 note 4} The only ‘important’ unsupported readings are those readings which, being apparently genuine, diverge too much from the rest to be due to accident, and those readings which reveal what the scribe had before him. To this latter class, of which I have spoken (on p. 35), belongs F's Tarpetia which, as Mr Housman says, indicates Tarpeiia in the exemplar.

^{page 181 note 5} This is why I have ‘refrained from collecting the similar evidence’againstΦ. Mr Housman also has refrained from collecting it.

^{page 182 note 1} Mr Housman errs in supposing I think D interpolated at II xxx 36 from I i 14; but I see now I did not make my meaning clear. The phrase (p. 37) ‘the scribe allowed himself to write’ (not ‘wrote’ or ‘interpolated’) I intended to be taken with the preceding, not the following context; and to exemplify the way in which the scribe allowed his mind to work upon his subject.

^{page 182 note 2} There is nothing against this but Baehrens'opinion that D's date is 1410–20. As Baehrens appears to have mis-dated all the Propertian manuscripts except F, whose first owners' names are written inside, this counts for nothing. As to V, which he places at the end of the 14th century, MM. Stevenson fils, Maurice Faucon, and Pierre de Nolhac agreethat this date is at least half a century ovit. And Mr E. Maunde Thompson's opinion, based on the photograph in Chatelain's Paléographie Latine, is that V was probably written ‘after the middle of the 15th century.’

^{page 183 note 1} The starveling theory then is his own, not that of Baehrens, on whom he would father it.

^{page 183 note 2} Mr Housman vapours here as follows: ‘The College of Cardinals rejected the simple account because it seemed to threaten Holy Writ. Dr Post-gate rejects the simple account because it is derogatory to the scarce le3S sacred Neapolitanus' I Thus does the habit of intemperance in language destroy the perception of proportion and even the sense of humour in its victim.

^{page 184 note 1} I think his date for A (1360) also somewhat too late. [Since the above was printed I have received Mr E. Maunde Thompson's opinion based on the page of A photographed in Chatelain's Paléographie Latine. He is ‘decidedly of opinion that the Vossianus was written about 1300 and just as likely before it as after it.’]

^{page 184 note 2} Mr Housman disputes the relevancy of this parallel because‘scribes are sometimes awake, and sometimes asleep.’ But there is a reason why the scribe should not have slept at II xxx 26. There was something the matter with the exemplar there; so the variants of Φ declare, ‘detcnete’ N¹, ‘tenete’ N²F, ‘tedere’ L, different combinations of the same elements, appear to indicate tcdcre in Φ, an error and its correction (as I have said p. 23 in a passage unfortunately misprinted). This the scribe of N misinterpreted, but the second hand corrected his mistake.

^{page 185 note 1} ‘Librarii erroies arguere ualebit e melioribns uulgaris notae libris quiciimque eligetur.’ Any suitably selected MS (this, not ‘any ’MS, appears to be the meaning of the Latin) of the better vulgar sort which would refute the errors of N, would have extrinsic validity comparable to that of the ‘unicus codex’ and to the world, if not to the writer, would be a sort of revelation.