When I was working on the Supplement to the Oxford English Dictionary, I was fascinated by a section tacked on to the end of the final volume of the original OED. It was entitled List of Spurious Words' and it was a collection of "the more important spurious words (arising chiefly from misprints or misreadings) that have been current in English dictionaries or other books of some authority".

Yes, even dictionaries make mistakes. Writing about Dr Johnson's famous dictionary, James Boswell wrote: "A few of his definitions must be admitted to be erroneous. Thus, Windward and Leeward, though directly of opposite meaning, are defined identically the same way towards the wind' . . . A lady once asked him how he came to define Pastern, the knee of a horse: instead of making an elaborate defence, as she expected, he at once answered, 'Ignorance, Madam, pure ignorance'."

Of course, Dr Johnson is notorious for some of his ironic or wordy definitions, like "Oats, a grain, which in England is generally given to horses, but in Scotland appears to support the people" or "Dross, the recrement or despumation of metals."

Johnson is well-known for his jaundiced view of the dictionary-writer, with his definition of lexicographer as "a harmless drudge". But he included a lesser-known jibe at lexicographers in this entry: "Grubstreet, originally the name of a street in Moorfields in London, much inhabited by writers of small histories, dictionaries, and temporary poems; whence any mean production is called grubstreet."

The OED's List of Spurious Words suggests that Johnson was not the only lexicographer whose definitions might arouse doubt or laughter. The OED's list includes Colophonian, which some dictionaries defined as relating to a colophon or the conclusion of a book' when it actually meant an inhabitant of Colophon, a town in Lydia, an ancient region of Asia Minor.

The word epidemic was mistakenly read as exidemic by some lexicographers, because of the similarity of the letters X and P in 16th and 17th century writing. Spurious words like these are often called ghost words and they can get into more than one dictionary, because lexicographers often borrow from one another.

One such word is abacot (a corruption of the word bycoket, meaning a kind of cap or head-dress), whose strange history is described by the OED thus: "Through a remarkable series of blunders and ignorant reproductions of error, this word appears in modern dictionaries as abacot. In Hall's Chron. a bicocket appears to have been misprinted abococket, which was copied by Grafton, altered by Holinshed to abococke, and finally improved' by Abraham Fleming to abacot (perhaps through an intermediate abacoc); hence it was again copied by Baker, inserted in his Glossarium by Spelman, and thence copied by Phillips, and so handed down through Bailey, Ash, Todd, etc., to 19th century dictionaries (some of which provide a picture of the abacot'), and even inserted in dictionaries of English and foreign languages."

The word dord was defined as density' in the second edition of Webster's New International Dictonary (1934) but it was a ghost word arising from a chemistry expert writing on a slip of paper "D or d, cont./density", meaning that D or d should be included as an abbreviation for density. The slip included cont' to indicate that more D abbreviations would be included. D or d' was taken to be a word, and thus dord got into the dictionary and it wasn't noticed for five years.

Non-existent words are sometimes deliberately inserted into dictionaries - either as a joke or to trap anyone illegitimately trying to copy the work. The first edition of the Collins COBUILD English Language Dictionary (1987) contained the fictitious word hink, defined thus: "If you hink, you think hopefully and unrealistically about something."

The New Oxford American Dictionary (2001) included the word esquivalience, which was defined as the wilful avoidance of one's official responsibilities.' When this dictionary entered a second edition in 2005, its Editor (Erin McKean) confessed that there was a non-existent word somewhere in the letter E. She eventually admitted that it was esquivalience, which had been entered to protect the copyright of the dictionary, both in book and electronic form.

This trick is also used in encyclopedias, where it is sometimes called a Mountweazel, after a fictitious entry for Lillian Virginia Mountweazel in the New Columbia Encyclopedia (1975). Lillian was purportedly an American photographer well-known for her series of photos of unusual subjects, including New York buses, the cemeteries of Paris, and rural American mailboxes (this last exhibited extensively abroad and published as Flags Up!). Her biography also says that she died in 1973 in an explosion while on an assignment for a magazine called Combustibles.

In 1903, Rupert Hughes's The Musical Guide ended one of its sections by listing a musical instrument called the zzxjoanw, which it said was a Maori word for a drum or a fife, pronounced shaw'. This was fairly obviously a hoax, especially if you knew that the Maori language contains no J, X or Z.

Tony Augarde is the author of The Oxford Guide to Word Games (OUP, £14.99) and The Oxford A to Z of Word Games (OUP, £4.99).