considering that the incomplete line here may be just as significant as an incomplete line ending a longer speech. If a speech begins within a line and ends brokenly, of course I have not counted it when it is equivalent to a five-foot line; e.g.
Wife, children, servants, all
That could be found:
but I do count such a speech (they are very rare) as
My lord, I do not know:
But truly I do fear it:
for the same reason that I count
You know not
Whether it was his wisdom or his fear.
Of the speeches thus counted, those which end somewhere within the line I find to be in Othello about 54 per cent.; in Hamlet about 57; in King Lear about 69; in Macbeth about 75.[285] The order is the same as König's, but the figures differ a good deal. I presume in the last three cases this comes from the difference in method; but I think König's figures for Othello cannot be right, for I have tried several methods and find that the result is in no case far from the result of my own, and I am almost inclined to conjecture that König's 41.4 is really the percentage of speeches ending with the close of a line, which would give 58.6 for the percentage of the broken-ended speeches.[286]
We shall find that other tests also would put Othello before Hamlet, though close to it. This may be due to 'accident'—i.e. a cause or causes unknown to us; but I have sometimes wondered whether the last revision of Hamlet may not have succeeded the composition of Othello. In this connection the following fact may be worth notice. It is well known that the differences of the Second Quarto of Hamlet from the First are much greater in the last three Acts than in the first two—so much so that the editors of the Cambridge Shakespeare suggested that [Q1] represents an old play, of which Shakespeare's rehandling had not then proceeded much beyond the Second Act, while [Q2] represents his later completed rehandling. If that were so, the composition of the last three Acts would be a good deal later than that of the first two (though of course the first two would be revised at the time of the composition of the last three). Now I find that the percentage of speeches ending with a broken line is about 50 for the first two Acts, but about 62 for the last three. It is lowest in the first Act, and in the first two scenes of it is less than 32. The percentage for the last two Acts is about 65.
II. The Enjambement or Overflow test is also known as the End-stopped and Run-on line test. A line may be called 'end-stopped' when the sense, as well as the metre, would naturally make one pause at its close; 'run-on' when the mere sense would lead one to pass to the next line without any pause.[287] This distinction is in a great majority of cases quite easy to draw: in others it is difficult. The reader cannot judge by rules of grammar, or by marks of punctuation (for there is a distinct pause at the end of many a line where most editors print no stop): he must trust his ear. And readers will differ, one making a distinct pause where another does not. This, however, does not matter greatly, so long as the reader is consistent; for the important point is not the precise number of run-on lines in a play, but the difference in this matter between one play and another. Thus one may disagree with König in his estimate of many instances, but one can see that he is consistent.
In Shakespeare's early plays, 'overflows' are rare. In the Comedy of Errors, for example, their percentage is 12.9 according to König[288] (who excludes rhymed lines and some others). In the generally admitted last plays they are comparatively frequent. Thus, according to König, the percentage in the Winter's Tale is 37.5, in the Tempest 41.5, in Antony 43.3, in Coriolanus 45.9, in Cymbeline 46, in the parts of Henry VIII. assigned by Spedding to Shakespeare 53.18. König's results for the four tragedies are as follows: Othello, 19.5; Hamlet, 23.1; King Lear, 29.3; Macbeth, 36.6; (Timon, the whole play, 32.5). Macbeth here again, therefore, stands decidedly last: indeed it stands near the first of the latest plays.