Fig. 77 is a design which was made by the writer for a flight of steps leading from a broad walk to a grass terrace about 2 feet 6 inches below. The building and terrace of which it was an accessory are very large, and it was imperative to keep the parts large in order to

avoid dwarfing the design and giving it a trumpery appearance. The balls are coincident with the centres of the pilasters in the wall at the back, which wall belongs to a large terrace in front of the house. style of the house is that prevalent in the reigns of George I. and George II.

The

With regard to flights of steps leading down to a level lower by only 3, 4, or 5 feet, or even more, there is no greater mistake than to use piers and balustrade

on the flight only. The flight is by no means more dangerous than the level, nor does it therefore require

FIG. 78.

a handrail in a greater degree. The writer has often seen the design treated as above, which will be seen to be a ridiculous method.

FIG. 79.

The above, fig. 79, will serve well for a raking balustrade, in connection with architecture of either the Elizabethan or Stuartian period. It is an original design of the writer's. More fancy in the vase would improve the effect.

Fig. 80 shows the usual way of designing a raking balustrade for the Elizabethan or Stuartian style of

architecture. In the writer's opinion the raking joints of beds and caps are wrong in principle and offensive

FIG. 80.

in appearance; and if the balusters are highly carved and decorated, as they frequently are, the difficulty of

making the raking mouldings look even decent is beyond conception, the members becoming alternately painfully acute and disagreeably obtuse. Compare fig. 82 with fig. 83, and it will be seen that fig. 82 must be square on

plan throughout, or it will

be distorted frightfully,

FIG. 82.

FIG. 83.

while fig. 83 may be either square or circular.

Fig. 81 is an example of the effect of keeping the bedding joints horizontal, there being not a single acute joint in the composition.

These designs admit of as much decoration as the means at the disposal of the professor will allow; but as they are to illustrate principles, the writer thought it well to make them as plain as possible.

The baluster, fig. 82, is not strictly classical, because in that style no one would think of making the members otherwise than horizontal.

It is never safe to allow a stone to be cut to less than a right angle, except in carving; that is to say, let no constructional or bedding joint be acute.

It will be seen that the coping at a, in example A, fig. 84, is cut to an acute angle: it is sure to look ill.

B

FIG. 84.

In the first place, the sharp corner a is nearly certain to break off; if it does not it will separate itself from the pier and present a very bad appearance. Example B shows the proper method. See b.

If the golden rule of never cutting a stone to less than a right angle were in all cases observed, architec

ture generally would be much improved. The following diagrams will show the bad effects of the neglect of this rule.

Fig. 85 shows two sorts of arches, or, rather, two

FIG. 85.

different methods of jointing. It will be seen that, excepting the key-stone and one stone on each side of it, there is not a single voussoir of the proper shape; those on the right, being cut in the same manner as is

cheese for retail purposes, must break at the acute angle a, a, a, fig. 86; while those on the left, not being