Medical texts comprise the Old English genre which attests most often to ælf. At the beginning of this book, I sketched the image extracted from this material in the early twentieth century, which characterised ælfe as small, mischievous spirits who caused illness by shooting arrows (a phenomenon called ‘elf-shot’). I have now also assembled the evidence for a quite different conception of ælfe: male, beautiful, human(-like), and otherworldly. It would be possible to square these conclusions with the medical texts simply by proposing that the medical texts exhibit the kind of demonisation of ælfe attested in Beowulf and the Royal Prayerbook. However, the need for a detailed and sustained reassessment of the medical texts, to see what evidence they really afford, is clear – and the resultant picture is both more complex and more interesting than what has hitherto been perceived. An important part of this revision has been done already: subsequently to her 1996 book, Jolly showed that the illustration to psalm 37 in the Eadwine Psalter, long imagined to depict ‘elf-shot’, is really a conventional depiction of demons, straightforwardly illustrating the psalm: ‘the later iconography of elves as delightfully mischievous little figures playing tricks on people has caused scholars such as Grattan and Singer to read an Anglo-Saxon elf into this picture of demonic affliction’. The present chapter focuses, then, purely on texts.

Three Anglo-Saxon medical manuscripts attest to ælf, usually in somewhat peripheral contexts, suggesting a certain ambivalence about the appropriateness of the material. I have discussed the late-tenth- or early-eleventh-century manuscript Harley 585 (in connection with Wið fǣrstice), and return to Wið fǣrstice at the end of this chapter; ælf occurs in Harley 585 once otherwise, in an attestation of ælfsīden considered in chapter 5. Likewise, I have discussed in chapter 2 the Royal Prayerbook's earlier, demonising attestation of ælf. Falling between these manuscipts in date is BL Royal 12 D. xvii, which contains the collections known as Bald's Leechbook (in two books) and Leechbook III. The manuscript is handsome if plain, written by the scribe who entered the batch of annals for 925–55 into the Parker Chronicle. This suggests that the manuscript was produced at Winchester in the mid-tenth century, the political bias of the Chronicle entries consolidating the obvious assumption of links to King Edmund's court.