ing of lines in Sonnet LV. are derived from a paffage in Meres's Palladis Tamia, 1598, where Shakspere among others is mentioned with honour:

'As Ovid faith of his worke;

Jamque opus exegi, quod nec Jovis ira, nec ignis, Nec poterit ferrum, nec edax abolere vetuftas ; And as Horace faith of his,

Exegi monumentum aere perennius,
Regalique fitu pyramidum altius;

Quod non imber edax, non Aquilo impotens
Poffit diruere, aut innumerabilis

Annorum feries et fuga temporum:

So fay I feverally of Sir Philip Sidney's, Spenfer's, Daniel's, Drayton's, Shakespeare's and Warner's workes ;

Nec Jovis ira, imbres, Mars, ferrum, flamma, fene&us, Hoc opus unda, lues, turbo, venena ruent.

Et quanquam ad pulcherrimum hoc opus evertendum tres illi Dii confpirabunt, Chronus, Vulcanus, et Pater ipfe gentis ;

Nec tamen annorum feries, non flamma, nec enfis, Aeternum potuit hoc abolere decus'.

III. The laft line of Sonnet XCIV.

Lilies that fefter fmell far worse than weeds

occurs also in the play King Edward III. (printed 1596), in a part of the play ascribed by fome critics to Shakspere. We cannot say for certain whether the play borrows from the sonnet, or the fonnet from the play. The latter feems to me the more likely supposition of the two.

The argument for this or that date from coincidences in expreffion between the Sonnets and certain plays of Shakspere has no decifive force. Coincidences may often be found between Shakfpere's late and early plays. But the general characteristics of style may lead us to believe that fome Sonnets, as I.-XXIV., belong to a period not later than Romeo & Juliet; others, as LXIV.-LXXIV., feem to echo the fadder tone heard in Hamlet and Measure for Measure. I cannot think that any of the Sonnets are earlier than Daniel's Delia' (1592), which, I believe, fupplied Shakspere with a model for this form of verfe; and, though I can allege no strong evi

dence for the opinion, I should not be difpofed to place any later than 1605.

Various attempts have been made by English, French, and German ftudents to place the Sonnets in a new and better order, of which attempts no two agree between themselves. That the Sonnets are not printed in the Quarto, 1609, at haphazard, is evident from the fact that the Envoy, CXXVI. is rightly placed; that poems addreffed to a mistress follow thofe addressed to a friend; and that the two Cupid and Dian Sonnets stand together at the close. A nearer view makes it apparent that in the first series, I.-CXXVI., a continuous story is conducted through various ftages to its termination; a more minute inspection discovers points of contact or connexion between fonnet and fonnet, and a natural fequence of thought, passion and imagery. We are in the end convinced that no arrangement which has been proposed is as good as that of the Quarto. But the force of this remark feems to me to apply with certainty only to Sonnets I.-CXXVI. The second series, cXXVII.-CLIV., al

though fome of its pieces are evidently connected with those which stand near them, does not exhibit a like intelligible fequence; a better arrangement may perhaps be found; or, it may be, no poffible arrangement can educe order out of the struggles between will and judgement, between blood and reafon; tumult and chaos are perhaps a portion of their life and being.

A piece of evidence confirming the opinion here advanced will be found in the use of thou and you by Shakspere as a mode of address to his friend. Why thou or you is chofen, is not always explicable; fometimes the choice feems to be determined by confiderations of euphony; fometimes of rhyme; fometimes intimate affection feems to indicate the use of you, and respecful homage that of thou; but this is by no means invariable. What I would call attention to, however, as exhibiting something like order and progress in the arrangement of 1609 is this: that in the first fifty fonnets, you is of extremely rare occurrence, in the second fifty you and thou alternate in little groups of fonnets,

thou having still a preponderance, but now only a flight preponderance; in the remaining twentyfix, you becomes the ordinary mode of addrefs, and thou the exception. In the fonnets to a mistress, thou is invariably employed. A few fonnets of the firft feries as LXIII.-LXVIII. have 'my love', and the third person throughout.1

Whether idealifing reality or wholly fanciful, an Elizabethan book of fonnets was-not always, but in many inftances-made up of a chain or series of poems, in a defigned or natural sequence, viewing in various aspects a single

1 I cannot here present detailed statistics. Thou and you are to be confidered only when addreffing friend or lover, not Time, the Muse, etc. Five fets of fonnets may then be distinguished: 1. Ufing thou. 2. Ufing you. 3. Ufing neither, but belonging to a thou group. 4. Ufing neither, but belonging to a you group. 5. Ufing both (xxiv.). I had hoped that this investigation was left to form one of my gleanings. But Profeffor Goedeke in the Deutsche Rundschau, March 1877, looked into the matter; his results seem to me vitiated by an arbitrary divifion of the fonnets ufing neither thou nor you into groups of eleven and twelve, and by a fantastic theory that Shakspere wrote his fonnets in books or groups of fourteen each.