My father once sent a friend a ‘short-copy’ of an article on the interaction of the Anglo-Saxon and the Celtic languages. A ‘short-copy’ is a reprint of such pages in a publication as make up an article; but the venerable man was not aware of this, and wrote back in a rage, Moreton, 28 July 1857, “Some confounded rascal has torn away 38 pages from the beginning of the work, and how many from the latter part I cannot say, it ends at 94. If I could get hold of the ears of the scoundrel, I would make them tingle. Such a gratuitous piece of mischief is enough to make a saint mad, for I dare say the fellow could make no more of his plunder than a pig.” In the English edition of Hanotaux’s _Contemporary France_ there is a footnote, vol. 1, page 127--“Demander a Bertrand le text Billet.” Bertrand was the librarian at the Ministry of Foreign Affairs; and I suppose Hanotaux wanted to quote the document at first hand from the archives instead of quoting it at second hand as printed in Lamberty. And his editor did not understand. The ancients also made mistakes like that. There is a Greek inscription at Stratoniceia, engraved in Fellows’ _Asia Minor_, pages 255, 256. It looks just like a table of accounts--the words begin at the left end of the lines, and at the right end there are figures, with an interval of varying length between the figures and the words. It really is a set of verses, and the figures give the number of letters in each line. They probably were meant to guide the mason in his work, but he has carved it all up on the temple wall. On the temple wall at Denderah there is a long inscription in Egyptian hieroglyphics with portraits of the deities who are mentioned in the text. At the portrait of Isis the hieroglyphics say “wood: gold: eyes of precious stone: two cubits high,” at the portrait of Heqit (the frog goddess) they say “silver overlaid with gold: five spans high,” and so on--it is engraved in Mariette’s _Denderah_, vol. IV, plate 88. The mason has carved his instructions on the wall, not knowing the meaning of the hieroglyphs. Goethe used often to dictate, but did not always look at what was written down; and years afterwards (_Werke_, vol. XLV, pp. 158 ff., ed. 1835) he found that Tugendfreund had been taken down as Kuchenfreund, John Hunter as Schon Hundert, and so on. Printers sometimes make mistakes. In a proof that I was reading for a friend, the nomen and pre-nomen of Rameses II had become his women and pre-women. But very often it is not the printer’s fault. In my former _Small Talk_ I wrote Anaxagoras for Aristagoras, and passed it in the proofs, page 76. Had it been anybody else’s work, I should have seen the error at once; but I suppose my mind was running on what I meant to say, and I thought I saw it there. And this may also be the reason why amateur artists very often fail to see the faults in their own work. They see the picture that they meant to paint, and not the daub that they have done.