We have yet to consider III. iv. 130-to end.

"And betimes I will to the weird sisters;"

the poverty of thought in

"For mine own good

All causes shall give way: I am in blood
Stept in so far, that, should I wade no more,
Returning were as tedious as go o'er :

Strange things I have in head that will to hand,

Which must be acted ere they will be scann'd;"

the putting this long tag in Macbeth's mouth when he is so bewildered that he answers Lady Macbeth's

by

"You lack the season of all natures, sleep

"Come, we'll to sleep,"

are all marks of inferior work, and make me sure that this part has been worked over by Middleton.

There is a passage in IV. i. that has been worked over in a similar way. After the speech of the third apparition Macbeth says,

"That will never be.

Who can impress the forest, bid the tree

Unfix his earth-bound root? Sweet bodements, good!
Rebellious dead, rise never till the wood

Of Birnam rise, and our high-placed Macbeth

Shall live the lease of nature, pay his breath

To time and mortal custom.'

"Our high-placed Macbeth cannot be said by Macbeth himself: it must be part of a speech of a witch. "Sweet bodements!" looks also like Middleton, and the whole bit is, in my opinion, a fragment of Hecate's inserted by him. "Rebellious dead" seems to me an allusion to Banquo's ghost, misplaced by Middleton. If we read "Rebellion's head" it seems a mistaken interpretation of the armedhead apparition: in any case, it is not Shakespeare. And I have no doubt a minute examination may detect still more traces of Middleton; but in an essay of this kind more detail would be wearisome.

Enough is given for my purpose to make it likely that Middleton was a recaster of the play, not a joint author.

Before giving my theory as to this play, and the metrical confirmations of it, I had better perhaps add a Table of the parts I do believe to be Shakespeare's.

I The part assigned by me to Middleton, but not by the Cambridge editors, is not 30 lines in all; I have asterised it in the table.-F. G. F.

This is an instance in which such editions as I have given in the Transactions of the New Shakspere Society of Marina (Pericles) and Timon would be worthless. Middleton certainly did not confine himself to adding to Shakespeare's work: he also re-modelled, re-wrote, and made large excisions. We ought to have an edition of this play in two types: the presumed alterations, and additions of Middleton's being in a smaller type than the rest, so that the better and more important portion might be read by itself.

I now give my theory as to the composition of the play. It was written by Shakespeare during his Third period : I think after Hamlet and Lear (see Malone); so that its date was probably 1606. Metrical evidence is of no use in determining the date: as we cannot tell how Middleton altered the play, or how much he omitted, except that the weak-ending test is not opposed to Malone's date. At some time after this, Middleton revised and abridged it: I agree with the Cambridge editors in saying not earlier than 1613. There is a decisive argument that he did so after he wrote the Witch, namely, that he borrows the songs from the latter play, and repeats himself a good deal. It is to me very likely that he should repeat himself in Macbeth, and somewhat improve on his original conception, as he has done in the corresponding passages and yet be unable to do a couple of new songs, or to avoid the monotony of introducing Hecate in both plays (Hecate being a witch in both, remember). I can quite understand a third-rate man, who in all his work shows reminiscences of others, and repetitions of Shakespeare, being unable to vary such conceptions as he had formed on the subject. I believe that Middleton, having found the groundlings more taken with the witches, and the cauldron, and the visions in IV. i. than with the grander art displayed in the Fate goddesses of I. iii., determined to amalgamate these, and to give us plenty of them. Hence the witches call themselves weird sisters in the lyric part of I. iii. hence the speech of Macbeth, "I will to-morrow to the weird sisters," &c. I believe also the extra fighting in the last scenes was inserted for the same reason. But finding that the magic and the singing and the fighting made the play too long-for a play of that kind cannot be endured to the length of an ordinary tragedy of Shakespeare's-he cut out large portions of the psychological Shakespeare work, in which, as far as quantity is concerned, this

play is very deficient compared with the three other masterpieces of world-poetry, and left us the torso we now have. That the taste of the mob is of the nature I assign to it, is evident enough from the way this play is put on the stage now. I am not play-goer enough to say how often it has been represented in my time without still further additions from Middleton's lyrics and Locke's music, but I think it cannot be very often. To hide the excisions, Middleton put on tags at the places where he made the scenes end: and to my thinking, if any one will compare the endings of the scenes where Shakespeare has left them without tags with those where I have tried to show that Middleton put them in, he will find that there is a great difference in the completeness of the scenes. Or try another experiment cut off the tags from the scenes where Shakespeare put them and those where Middleton put them; a similarly decisive result will be felt. It is impossible to show this in a paper if I were doing an edition of the play with the opportunity of summing up the æsthetic of each scene at the end of it as I went on, I am certain I could make it manifest: not to mention many smaller details I cannot stay to discuss here, such as the stage direction in IV. i. about Banquo's carrying the glass. But I must stay to protest against the modern way of altering and inserting stage directions ad libitum; it has thrown back our criticism twenty years. I could not myself stir in this matter till I obtained reprints of Folio and Quartos, which I could not for many years, for reasons I need not dwell on here. I do not think we should do well in issuing mere reprints only, but no alteration even in popular editions should be made without being marked by brackets or italics, or some warning that there is an alteration: unless in correction of mere printers' errors, or in arranging the lines, or in punctuation.

From the nature of the

We now come to the metrical evidence. interpolations in the rhymes, &c., our usual tests are not attainable. Fortunately there are others that are. I give first, then,

From this table we see that all the last eight plays fall into two classes. One class consists of three early plays which were produced before Shakespeare had learnt his work as a playwright, however much he excelled already as a poet. The other is composed altered by some other

of five plays, four of which were finished or poet, as I have myself tried to show, and Mr. Staunton has satisfactorily accounted for the fifth (The Tempest). It cannot be accident that five plays1 thus altered should fall among the eight shortest of the total series of 38.

The chance of such an event is I in 8962: there must be a cause. One possible cause is assignable. We know from comparing the Quartos and Folio of Lear, Hamlet, Othello, Richard III., &c., that

The two plays finished by Fletcher do not fall under this category. They are of Fletcher's average length.