aw, but it does save time, and produces
better drawings than can be made without it. The best place to buy them
is of Cevalier, on the Seine, near the Pont Netif, Paris. Only those
with the best prisms are of any use: such a one, with two adjustments
only, can be had for sixty-five francs. The table which is necessary for
its use costs fifteen francs additional; that is, a total cost of
sixteen dollars. In buying a table, be sure and get one with sliding
legs which can be taken off the head and packed flat.

"One of the very best ways to study, and one which has very direct
tangible results, is by the aid of printed plates. Take such a book as
Letarouilly's _Edifices de Rome Moderne_. Go to the buildings themselves
and compare the drawing with the building; see what drawings on paper
really mean when executed; mark up the plate; note the proportion of
masses, the size of ornament, the relative proportion of openings, and
wall spaces, the effect of color and texture, and the use of material.
Make suggestions for better ornament, proportion, etc., and then go home
and make a new design with all the improvements you have noted.

"The reverse of this method is, to sit down in front of the building
with T-square and triangle and translate the perspective building back
on to paper in elevation.

[Illustration: XIV. Detail of the Principal Doorway to the Basilica at
Altamura, Italy.]

"These two methods will aid one to tell from a drawing how the building
will actually look when executed. It will give an idea of the scale of
ornament, if a cornice looks just the right size on a certain building,
the plate will tell you just how high that is. The T-square sketch is
very valuable in cultivating the sense of proportion. Draw to scale such
parts of the sketch as can be easily measured, and put in the remainder
in proportion, and make these sketches at the scale at which you are
used to working in the office. They will be of immense advantage in
giving you a sense of absolute scale.

"There is such a thing as 'absolute scale,' and scale is not simply
proportion. A drawing might be made in good proportion, and the building
look well if executed a thousand feet long, and yet lose all its
effectiveness if executed but one hundred feet in length, the relative
proportions of the parts remaining the same. It is a fact that certain
designs, which look well on paper, will not look well in execution,
except at a large scale. Therefore