in large red letters--Roman capitals for
choice--the name of the work: and it need hardly be said that these
should be arranged in alphabetical order. For minor matters too small
for such classification it is well to have, in the _front_ place in the
basket, cards dividing the alphabet itself into about four parts, so
that unarranged small matters can be still kept roughly alphabetical.
When the work is done, transfer all documents to separate labelled
portfolios--a folded sheet of the thickest brown paper, such as they put
under carpets, is very good--and store them away for reference. Larger
portfolios for all _templates_, tracings, or architects' details or
drawings relating to the work. If you have not a good system with regard
to the ordering of these things, believe me the mere _administration_ of
a very moderate amount of work will take you _all your day_.

So also with _measurement_.

ON ACCURACY IN MEASUREMENT.

In one of Turgenieff's novels a Russian country proverb is
quoted--"Measure thrice, cut once." It is a golden rule, and should be
inscribed in the heart of every worker, and I will add one that springs
out of it--"Never trust a measurement unless it has been made by
yourself, or for yourself--to your order."

The measurements on architects' designs, or even working drawings, can
never be trusted for the dimensions of the built work. Even the
builders' templates, by which the work was built, cannot be, for the
masons knock these quite enough out, in actual building, to make your
work done by these guides a misfit. Have your own measurements taken
again. Above all, beware of trusting to the supposed verticals or
horizontals in built work, especially in tracery. A thing may be
theoretically and intentionally at a certain angle, but actually at
quite a different one. If level is important, take it yourself with
spirit-level and plumb-line.

With regard to accuracy of work _in the shop_, where it depends on
yourself and the system you observe, I cannot do better than write out
for you here the written notice by which the matter is regulated in my
own practice with regard to cartoons.

_"Rules to be Observed in Setting out Forms for Cartoons._

"In every case of setting out any form, or batch of forms, for new
windows the truth of the first long line ruled must be _tested_ by
stretching a thread.

If the lath is proved to be out, it must at once be sent to a joiner to
be accurately 'shot,' and th