The second Beowulf scribe often begins the letter <e> with a thick downstroke as for the letter <i>, to which he adds a head and tongue sometimes thinner than the downstroke. On some damaged folios it is debatable whether a certain letter should be read as <i> itself or the remains of <e> in which the head and tongue of <e> are very obscure or simply gone. Further, on fols. 179 and 198v, a later hand, attempting to retrace obscure letters, occasionally restored a letter as <i> that the palaeographic evidence hints was once <e>. At other places the later hand restored a letter as <i> for which there is no evidence for <e> but where <e> is expected. Although the question of <i> for <e> and <e> for <i> is complex, careful study of the comparative evidence, beginning with the eighteenth-century transcripts, A and B, allows us to reach reasonable conclusions.