Although every possible care has been taken in selecting the best authorities for the statements in the text of the work, as well as the subjects for illustration, still no one acquainted with the state of the literature of architecture will need to be told that in many branches few materials exist for a correct description of the style, and that the drawings which are available are frequently so inexact and with scales so carelessly applied, that it is impossible at times to avoid error. The plans throughout the book are on too small a scale to render any minute errors apparent, but being drawn to a uniform scale of 100 feet to 1 inch, or one twelve-hundredth of the real size, they are quite sufficient as a means of comparison, even when not mathematically correct. They suffice to enable the reader to judge of the relative size of two buildings by a mere inspection of the plans, as correctly as he could by seeing the buildings themselves, without actually measuring them in all their details.
As a general rule, the sections or elevations of buildings, throughout the book, are drawn to a scale double that of the plans, viz., 50 feet to 1 inch, or one six-hundredth of the real dimensions; but, owing to the great size of many of them, it has been found impossible to carry out this in all instances: where it has not been effected the departure from the rule is always noted, either below the woodcut or in the text.
No lineal dimensions are quoted in the text except such as it is believed can be relied upon, and in all instances these are reduced to English feet. The superficial measures also in the text, like the plans, are quite sufficient for comparison, though not to be relied upon as absolutely correct. One great source of uncertainty as regards them is the difficulty of knowing at times what should be included in the building referred to. Should, for instance, the Lady Chapel at Ely be considered an integral part of the Cathedral, or the Chapel-house at Wells? Should the sacristies attached to Continental cathedrals be considered as part of the church? or such semi-detached towers as the south-western one at Bourges? What constitutes the temple at Karnac, and how much of this belongs to the Hypostyle Hall? These and fifty other questions occur in almost every instance which may lead two persons to very different conclusions regarding the superficial dimensions of a building, even without the errors inherent in imperfect materials.
When either the drawing from which the woodcut is taken was without a scale, or the scale given could not be depended upon, "No scale" has been put under the woodcut, to warn the reader of the fact. When the woodcut was either too large for the page or too small to be distinct if reduced to the usual scale, a scale of feet has been added under it, to show that it is an exception to the rule.
Capitals, windows, and details which are meant to illustrate forms or construction, and not particular buildings, are drawn to any scale that seemed best to express the purpose for which they are inserted; but when they are remarkable for size, or as individual examples, a scale has been added; but this is the exception, not the rule.
Every pains has been taken to secure the greatest possible amount of accuracy, and in all instances the sources from which the woodcuts have been taken are indicated. Many of the illustrations are from original drawings, and of buildings never before published.