§ =165.= In the first place the numerical proportion of the rhymed and rhymeless lines in a play deserves attention. Blank verse, it is true, prevails in all Shakespeare's plays; but in his undoubtedly earlier plays we find a very large proportion of rhymed verse, while in the later plays the proportion becomes very small. Some statistical examples, based on careful researches by English and German scholars, may be quoted to prove this; for the rest we refer to the special investigations themselves. In _Love's Labour's Lost_, one of Shakespeare's earliest plays, we have 1028 rhymed lines and 579 unrhymed. In _The Tempest_, one of his last plays, we find 1458 unrhymed and only two rhymed five-foot lines. In the plays that lie between the dates of these two dramas the proportion of rhymed and unrhymed verse lies between these two numbers. In _Romeo and Juliet_, e.g. (which belongs to the end of Shakespeare's first period, though Fleay thought it a very early play) we have 2111 unrhymed and 486 rhymed five-foot lines; in _Hamlet_ (belonging to the third period) there are 2490 unrhymed and 81 rhymed lines. In many cases, however, the use of rhyme in a play is connected with its whole tone and character, or with that of certain scenes in it. The frequency of rhymes in _Romeo and Juliet_ finds its explanation in the lyrical character of this play. For the same reason _A Midsummer Night's Dream_, although it is certainly later than _Love's Labour's Lost_ and _Romeo and Juliet_, shows a larger proportion of rhymed lines (878 blank: 731 rhymes). This seems sufficient to show that we cannot rely exclusively on the statistical proportion of rhymed and unrhymed verses in the different plays in order to determine their chronological order. § =166.= The numerical proportion of feminine and masculine endings is of similar value. In the early plays we find both masculine and feminine endings; the masculine, however, prevail. The number of feminine endings increases in the later plays. On this point Hertzberg has made accurate statistical researches. According to him the proportion of feminine to masculine endings is as follows: _Love's Labour's Lost_ 4 per cent., _Romeo and Juliet_ 7 per cent., _Richard III_ 18 per cent., _Hamlet_ 25 per cent., _Henry VIII_ 45·6 per cent.[159] This proportion, however, as has been shown by later inquiries,[160] does not depend solely on the date of the composition, but also on the contents and the tone of the diction, lines with masculine endings prevailing in pathetic passages, and feminine endings in unemotional dialogue, but also in passionate scenes, in disputations, questions, &c.