Many sets of ivory (or bone) alphabet letters have survived in private and institutional collections: the British Library has the one Jane Austen and her family used for playing word games. Perhaps they were taken out to solve riddles.
Here is a classic one Austen almost certainly would have known, because it was reprinted so often of eighteenth- and nineteenth-century collections of word games, usually without credit to its ingenious creator Jonathan Swift, the author of Gulliver’s Travels.
We are little airy creatures,
All of diff’rent voice and features;
One of us in glass is set,
One of us you’ll find in jet,
One of us is set in tin,
And the fourth a box within;
If the last you should pursue,
They can never fly from you.

[Alphabet of Bone Letters]. [England, 183-?]. (Cotsen 3507)
Perhaps customers were been offered a choice of boxes at point of sale. The Puzzle Museum owns a carved bone box containing bone letters with a sliding lid about the same size and shape as Cotsen’s, but with different decorations.
The distinctive letter forms of the bone ones, however, offer some evidence for dating them. They are clearly copies of the early nineteenth-century Roman “fat faces” made popular by Robert Thorne, Vincent Figgins, and William Thorowgood, that are the forerunners of slab serif and typewriter fonts. Here’s an early nineteenth-century handbill that uses one of these wonderful in-your-face fonts that were designed for use in advertising.
And here are the words “Turnip Seeds” spelled out in the Cotsen set of bone letters.
The resemblance is unmistakable, even though the bone letters’ serifs are not as skinny and spiky as in the fat faces and the contrasts between the thick and thin strokes not as exaggerated. If the bone letters had been more faithful copies of the typeforms, they probably would have been much too fragile to hold up to extended play. Here’s second comparison of type and bone letter:
And here is ” Durham” spelled out in Cotsen’s bone letters:
While redescribing Cotsen’s collection of alphabet tiles and letters this summer, I noticed that all but this set used some variation of the slab serif, instead of the Figgins fat face. This could mean that those alphabet letters can’t possibly date before 1815 or so, when the faces they were modeled after began to appear in type specimens and in job printing. So much for the common misconception that they are relics of eighteenth-century nursery artifacts…. In fact, I strongly suspect most of Cotsen’s sets could date to the mid- to late nineteenth century, but that’s a riddle for another time.