from the opening lines and the last act, winding up as it does with Puck's "Epithalamium," it is not improbable that it was, at least eventually, intended for the celebration of the marriage of some nobleman of Elizabeth's court; but I rather incline to the belief that it was not so in the first instance; and that, marriage or no marriage, we should have had A Midsummer-Night's Dream, though, perhaps, not exactly in its present form. "If," says Furness, "a noble marriage before 1598 can be found to which there are unmistakeable allusions in the play, we shall go far to confining the Date of Composition within narrow limits." Various attempts have been made to discover the marriage in question. The suggestion of Fleay is, in my opinion, by far the most probable yet made. In his Life and Work of Shakespeare (1886, p. 81), he says: "January 26 was the date of the marriage of William Stanley, Earl of Derby, at Greenwich. Such events were usually celebrated with the accompaniment of plays or interludes, masques written specially for the occasion not having yet become fashionable. The company of players employed at these nuptials would certainly be the Chamberlain's (ie. the company to which Shakespeare belonged), who had, so lately as the year before (ie. 1594) been in the employ of the Earl's brother Ferdinand. No play known to us is so fit for the purpose as A Midsummer-Night's Dream, which in its present form is certainly of this date. About the same time Edward Russel, Earl of Bedford, married Lucy Harington. Both marriages may have been enlivened by this performance. The date of the play here given is again confirmed by the description of the weather (in II. i. 81 sqq.). . . Chute's Cephalus and Procris was entered on the Stationers'

Registers, 28 Sept. 1593; Marlowe's Hero and Leander, 22nd October 1593; Marlowe and Nash's Dido was printed in 1594. All these stories are alluded to in the play. The date of the Court performance must be in the winter of 1594-95."

Marriage is the theme of the play. It is initiated by the coming marriage of Theseus and Hippolyta, and it is wound up not only by their marriage, but by those of the pairs of lovers. If Fleay's hypothesis be correct, may not this have some slight reference to the double wedding of 1594-95 ?

William Stanley was the younger brother of Ferdinand, Lord Strange, and by the death of his father in September 1593, and of his brother Ferdinand in April 1594, he became sixth Earl of Derby. Next year he married Elizabeth Vere, daughter of the Earl of Oxford, and Stowe, in his Annals, thus records the event:

"The 26 of January William Earl of Derby married the Earl of Oxford's daughter at the Court then at Greenwich, which marriage feast was there most royally kept."

It may, therefore, with some reason be conjectured, but only conjectured, that Elizabeth herself was present, and that the royal ears listened to the graceful though somewhat irrelevant tribute to the "fair vestal throned by the west" (II. i. 158). Inasmuch as the "marriage feast was most royally kept," in all likelihood one of the entertainments was A Midsummer-Night's Dream. Again, on the assumption that the play was performed at Greenwich and at William Stanley's wedding, it is not a further

stretch of probability to assume the presence of the Dowager Countess of Derby, the widow of the late Earl, who was Lady Strange at the date of the dedication to her of Spenser's poem in 1591; or, further, to assume that the reference to "the thrice three Muses" may have been intended as a compliment to her and the Stanley family. We must never forget, however, that in these matters we are forced, from the very circumstances, to deal with probabilities and not with actual facts; and it must also be noted that Shakespeare's company is not stated to have played at Court on "the 26 of January," though performances are recorded on the 5th January and 22nd February 1595. (See Fleay, Life and Work of Shakespeare, 1866, pp. 126, 127.)

4. Another allusion is distinctly in favour of the autumn of 1594. The reference in I. ii. 77 and III. i. 31 to the lion frightening the duchess and the ladies, is not improbably a reminiscence of an incident which happened at the Scottish Court at the baptism of Prince Henry, the eldest son of James I., in August 1594. Malone was the first to remark on "the odd coincidence," as he calls it. He quotes a pamphlet which is reprinted in Somers's Tracts, ii. 179: "While the king and queen were at dinner a chariot was drawn in by'a black-moore. This chariot should have been drawne in by a lyon, but because his presence might have brought some fears to the nearest, or that the sights of the lights and the torches might have commoved his tameness, it was thought meete that the Moor should supply that room.'"

Steevens, in his note to II. i. 15, refers to the following passage from the old anonymous comedy of The Wisdom

of Doctor Dodypoll, the earliest known edition of which is dated 1600:

'Twas I that led you through the painted meads,
When the light fairies danc'd upon the flowers,
Hanging on every leaf an orient pearl.

It is true that Nash, in his preface to Gabriel Harvey's Hunt is Up, 1596, mentions the name "doctor Dodypowle," but this is without any reference to the play, and the name Dodipoll had long previously been in use for a blockhead. H. Chichester Hart points out (Athenæum, 1888) that the name occurs in Hickscorner, 1552:

What, Master Doctor Dotypoll,

Cannot you preach well in a black boll
Or dispute any divinity?

It seems to be represented by the modern slang word "dotty." So that we can deduce no argument as to the date from the reference by Steevens.

The

The arguments of Chalmers for assigning the date of the play to the year 1598 may be found set out at length in Furness's New Variorum, p. 248 sqq. I shall not attempt to introduce them here, as, in my opinion, they have no real weight, and are weak and inconclusive. conjectures of Gerald Massey and of some of the German critics (Tieck, Elze, Kurz, and others), which attempt to fix an earlier or later date for the play, on the theory, amongst others, that the occasion of the performance was the marriage of Lord Essex with Lady Sidney in 1590, or that of Lord Southampton with Elizabeth Vernon in 1598-both secret marriages, by the way, and obnoxious to the Queen's displeasure-may also be found duly set

out in Furness, p. 248 sqq. In my opinion, they may be dismissed as not worth serious discussion.

A somewhat shrewd line of argument as to the date has been adopted by Aldis Wright in his Introduction to the Clarendon Press Edition, p. xi, where he says: "If we attempt to arrange the plays which Meres attributes to Shakespeare, so as to distribute them over the period from 1589 to 1598, we shall find two gaps, in either of which we might conjecturally place the Midsummer-Night's Dream. The interval from 1589 to 1591 is filled up by Love's Labour's Lost, the Two Gentlemen of Verona, Comedy of Errors, and Titus Andronicus. In 1593, 1594, are placed Richard the Second, Richard the Third, King John, and in these years appeared Venus and Adonis and Lucrece. The Merchant of Venice is assigned to 1596, and Henry the Fourth to 1597. Besides these there are the three Parts of Henry the Sixth, which Meres does not mention, but which, if Shakespeare's at all, must belong to the earlier part of this period, and Loue Labours Wonne,'

On the whole, I am dis

whatever this may have been. posed to agree with Professor Dowden in regarding the Two Gentlemen of Verona as earlier than the MidsummerNight's Dream, while I cannot think the latter was composed after the plays assigned above to 1593, 1594, and would therefore place it in the interval from 1591 to 1593, when perhaps Romeo and Juliet may have been begun.” I see no reason whatever to think that the historical plays above mentioned, i.e. those assigned to 1503, 1594, were necessarily composed after the Midsummer-Night's Dream. On the contrary, I am strongly of opinion that these historical plays show clearly that Shakespeare was still more