which corrections he has to pay at so much per
hour--second cause of war.

The proofs are now returned to the printer and corrected, and a revise
(after passing through the proof-reader's hands) sent to the author,
which process may be repeated _ad infinitum_, until the author gives
the order to make up into pages.

The type is now handed over to the "make-up," and inasmuch as his work
must be carefully revised by the proof-reader, we may describe it
here. Having first of all made a gauge showing the size of the
page--supposing the page to be seven inches deep, he will cut a notch
in a thin piece of wood showing that size--he must "cast off" or
estimate how the pages are going to "break." There must not be any
short lines, or "widows" as the printers call them,--that is, the
concluding lines of paragraphs which are not full length,--at the
heads of pages. The first line of a paragraph should not appear at the
bottom of a page (but this rule is more honored in the breach than the
observance), and the concluding page of a chapter should not be less
than one-quarter page in length. These difficulties are avoided by
"saving" a line here and there,--that is, where the last line of a
paragraph consists of only one or two words, in squeezing them into
the line above, or by "making" lines, which is accomplished by
spreading long lines out and driving one or two words over. Any line
containing one word only at the end of a paragraph ought to overlap
the indention of the first line of the next paragraph. Such a word as
"is" or "it" will not do so and should be turned back to the line
above. Then again, where cuts or illustrations are inserted in the
text a page will sometimes break in the middle of a cut, which, as
Euclid says, is impossible, therefore the cut must be moved, sometimes
necessitating slight alterations in the text, _e.g._ "The following
illustration" must be altered to "The illustration on the next page,"
or "The illustration above," as the case may be. And here we may
remark that all cuts or illustrations should be made and furnished to
the printer in time to be inserted in the first proof. The writer
calls to mind an instance where the cuts arrived after the whole book
had been made up into pages, necessitating a re-make-up at
considerable expense.

Proofs of the pages being furnished to the proof-reader, he first of
all compares them with the author's last galley proof to see that
nothing has been omitted (