Now the difficulty of the art was overcome in its earliest and best days, and no care or pains were spared in making the most of its special qualities, while for long and long no force was put upon the material to make it imitate the qualities of brush-painting, either in power of colour, in delicacy of gradation, or intricacy of treating a subject; and, moreover, easy as it would have been to minimise the jointing of the tesserae, no attempt was made at it.

But as time went on, men began to tire of the solemn simplicity of the art, and began to aim at making it keep pace with the growing complexity of picture painting, and, though still beautiful, it lost colour without gaining form.  From that point (say about 1460), it went on from bad to worse, till at last men were set to work in it merely because it was an intractable material in which to imitate oil-painting, and by this time it was fallen from being a master art, the crowning beauty of the most solemn buildings, to being a mere tax on the craftsmen’s patience, and a toy for people who no longer cared for art.  And just such a history may be told of every art that deals with special material.

Under this head of order should be included something about the structure of patterns, but time for dealing with such an intricate question obviously fails me; so I will but note that, whereas it has been said that a recurring pattern should be constructed on a geometrical basis, it is clear that it cannot be constructed otherwise; only the structure may be more or less masked, and some designers take a great deal of pains to do so.

I cannot say that I think this always necessary.  It may be so when the pattern is on a very small scale, and meant to attract but little attention.  But it is sometimes the reverse of desirable in large and important patterns, and, to my mind, all noble patterns should at least look large.  Some of the finest and pleasantest of these show their geometrical structure clearly enough; and if the lines of them grow strongly and flow gracefully, I think they are decidedly helped by their structure not being elaborately concealed.

At the same time in all patterns which are meant to fill the eye and satisfy the mind, there should be a certain mystery.  We should not be able to read the whole thing at once, nor desire to do so, nor be impelled by that desire to go on tracing line after line to find out how the pattern is made, and I think that the obvious presence of a geometrical order, if it be, as it should be, beautiful, tends towards this end, and prevents our feeling restless over a pattern.

That every line in a pattern should have its due growth, and be traceable to its beginning, this, which you have doubtless heard before, is undoubtedly essential to the finest pattern work; equally so is it that no stem should be so far from its parent stock as to look weak or wavering.  Mutual support and unceasing progress distinguish real and natural order from its mockery, pedantic tyranny.