92.  But this latter piece of criticism seems open to one grave objection to which the former is not liable.  Mr. Fleay separates the portions of the play which are undoubtedly to be assigned to witches from the parts he gives to his Norns, and attributes them to different characters; the other mixes up the witch and Norn elements in one confused mass.  The earlier critic saw the absurdity of such a supposition when he wrote:  “Shakspere may have raised the wizard and witches of the latter parts of Holinshed to the weird sisters of the former parts, but the converse process is impossible."[1] Is it conceivable that Shakspere, who, as most people admit, was a man of some poetic feeling, being in possession of the beautiful Norn-legend—­the silent Fate-goddesses sitting at the foot of Igdrasil, the mysterious tree of human existence, and watering its roots with water from the sacred spring—­could, ruthlessly and without cause, mar the charm of the legend by the gratuitous introduction of the gross and primarily unpoetical details incident to the practice of witchcraft?  No man with a glimmer of poetry in his soul will imagine it for a moment.  The separation of characters is more credible than this; but if that theory can be shown to be unfounded, there is no improbability in supposing that Shakspere, finding that the question of witchcraft was, in consequence of events that had taken place not long before the time of the production of “Macbeth,” absorbing the attention of all men, from king to peasant, should set himself to deal with such a popular subject, and, by the magic of his art, so raise it out of its degradation into the region of poetry, that men should wonder and say, “Can this be witchcraft indeed?”

[Footnote 1:  Shakspere Manual, p. 249.]

93.  In comparing the evidence to be deduced from the contemporary records of witchcraft with the sayings and doings of the sisters in “Macbeth,” those parts of the play will first be dealt with upon which no doubt as to their genuineness has ever been cast, and which are asserted to be solely applicable to Norns.  If it can be shown that these describe witches rather than Norns, the position that Shakspere intentionally substituted witches for the “goddesses of Destinie” mentioned in his authority is practically unassailable.  First, then, it is asserted that the description of the appearance of the sisters given by Banquo applies to Norns rather than witches—­

    “They look not like the inhabitants o’ th’ earth,
    And yet are on’t.”

This question of applicability, however, must not be decided by the consideration of a single sentence, but of the whole passage from which it is extracted; and, whilst considering it, it should be carefully borne in mind that it occurs immediately before those lines which are chiefly relied upon as proving the identity of the sisters with Urda, Verdandi, and Skulda.

Banquo, on seeing the sisters, says—­