It must therefore be carefully borne in mind, that this edition of 1845, on the title-page of which the name of Tyrwhitt is so conspicuous, was really edited anonymously, or may even be said not to have been edited at all. The Canterbury Tales are reprinted from Tyrwhitt; and so also are the Essay, the Discourse, the Notes, and the Glossary; and it is most important to observe that 'the Glossary' is preceded by Tyrwhitt's 'Advertisement,' and by his 'Account of the Works of Chaucer to which this Glossary is adapted; and of those other pieces[7] which have been improperly intermixed with his in the Editions.' The volume is, in fact, made up in this way. Pages i-lxx and 1-209 are all due to Tyrwhitt; and contain a Preface, an Appendix to the Preface, an Abstract of Passages of the Life of Chaucer, an Essay, an Introductory Discourse to the Tales, and the Tales themselves. [xiv] Again, pp. 441-502 are all due to Tyrwhitt, and contain an Advertisement to the Glossary, an Account of Chaucer's Works (as above), and a Glossary. Moreover, this Glossary contains a large number of words from most of Chaucer's Works, including even his _prose_ treatises; besides a handful of words from spurious works such as 'Chaucer's Dream.'
In this way, all the former part and all the latter part of the volume are due to Tyrwhitt; it is the _middle part_ that is wholly independent of him. It is here that we find no less than twenty-five poems, _which he never edited_, reprinted (inexactly) from the old black-letter editions or from Chalmers. It thus becomes plain that the words 'By Thomas Tyrwhitt' on the title-page refer only to the second clause of it, but have no reference to the former clause, consisting of the words, 'The Poetical Works of Geoffrey Chaucer.' It remains to be said that the twenty-five poems which are here appended to the Canterbury Tales are well selected; and that the anonymous editor or superintendent was guided in his choice by Tyrwhitt's 'Account of the Works.'
§ 8. This somewhat tedious account is absolutely necessary, every word of it, in order to enable the reader to understand what has always been meant (since 1845) by critics who talk about some works as being 'attributed to Chaucer.' They really mean (in the case, for example, of The Cuckoo and the Nightingale) that it happens to be included in a certain volume _by an anonymous editor_, published in 1845, in which the suggestions made by Tyrwhitt in 1778 were practically adopted without any important deviation. In the case of any other author, such a basis for a canon would be considered rather a sandy one; it derives its whole value from the fact that Tyrwhitt was an excellent literary critic, who may well be excused for a few mistakes, considering how much service he did in thus reducing the number of poems in 'Chaucer's Works' from 64 to little more than 26[8]. Really, this was a grand achievement, especially as it clearly emphasised the absurdity of trusting to the old editions. But it is an abuse of language to say that 'The Cuckoo and Nightingale' has 'always been attributed to Chaucer,' merely [xv] because it happens to have been printed by Thynne in 1532, and had the good luck to be accepted by Tyrwhitt in 1778. On the contrary, such a piece remains on its trial; and it must be rejected absolutely, both on the external and on the internal evidence. Externally, because no scribe or early writer connects it with him in any way. Internally, for reasons given in vol. i. p. 39[9]; and for other reasons given in Lounsbury's Studies in Chaucer.