But inasmuch as there were no shields actually in use, the shapes prevailing now became entirely a matter of fashion and taste. Randle Holme prints sixty-five varieties of shields, and remarks (p. 10 in his _Academy of Armoury_, Chester, 1688):--"But [as to the former shield] so to this, a question may be made, whether such an one was used by him, or only the invention of the cutter? If so, then the shapes and forms of shields, targets, and bucklers would be as many as carvers, stone cutters, engravers, and painters please." Gerard Legh gets the credit of having greatly multiplied, in his _Accedens of Armoury_, 1562, the fanciful shapes of shields; and later writers on heraldry seem to have followed him. [_Herald and Genealogist_, vol. i, p. 191.] But an examination of the _Wappenbüch_, published in Germany in the sixteenth and seventeenth centuries, and similar heraldic drawings of Italian coats, have opened out to me a vastly wider field for variety than old Randle Holme ever supposed. I have before me a copy of _Virgil Solis_, which was published at Nurnberg in 1555. This contains hundreds of coats, in shields of an endless variety of shapes and great boldness of design; and when we come to critically compare these shields with one another, we discover that their outlines depend upon the charges! These were _first_ laid down with grotesque boldness, and the outline of the shield to bear them was then drawn around them, to the artist's fancy of what best suited the bearings and would most reduce the amount of empty field. This is, after all, only reversing the ancient method, where a lion was distorted to suit a pointed heater; and it is suggested that a pale, for a similar reason, might sometimes be pointed into a pile. With this revelation from _Virgil Solis_, we can detect that the same method of designing prevails elsewhere. We must therefore multiply Randle Holme's carvers, stone cutters, engravers, and painters by the possible variety of their fancies, and then multiply the quotient by the number of all the heraldic bearings--plus all the positions and forms they may be made to assume--if we wish to arrive at the variety of shields possible during the period we are now about to consider. In deference to this unanswerable position, I find it necessary to alter entirely the classification pursued in the earlier centuries--to take Mr. Rylands' drawings seriatim, and to give for each shape the earliest and latest dates at which, in my limited researches, I have found them in use. This method of treatment eliminates from our enquiry all strictly fanciful shields which did not gain a hold on public taste, and leaves us still with a sufficiently heavy list of variations to which it is possible, from their frequent recurrence, to attach the dates. A few leading remarks on points bearing on our subject will prove useful at this stage of our enquiry.