he has selected his contributor, feels that {17} the
contributor is likely to know his business better than an editor can teach
him; in fact, it is on that principle that the selection is made. But he
feels that he is more competent than the writer to judge questions of
strength and of tone, especially when the general purpose of the journal is
considered, of which the editor is the judge without appeal. An editor who
meddles with substantive matter is likely to be wrong, even when he knows
the subject; but one who prunes what he deems excess, is likely to be
right, even when he does not know the subject. In the second place, a
contributor knows that he is supplying an editor, and learns, without
suppressing truth or suggesting falsehood, to make the tone of his
communications suit the periodical in which they are to appear. Hence it
very often arises that a reviewed author, who thinks he knows the name of
his reviewer, and proclaims it with expressions of dissatisfaction, is only
wrong in supposing that his critic has given all his mind. It has happened
to myself more than once, to be announced as the author of articles which I
could not have signed, because they did not go far enough to warrant my
affixing my name to them as to a sufficient expression of my own opinion.

"There are two other ways in which a reviewed author may be wrong about his
critic. An editor frequently makes slight insertions or omissions--I mean
slight in quantity of type--as he goes over the last proof; this he does in
a comparative hurry, and it may chance that he does not know the full sting
of his little alteration. The very bit which the writer of the book most
complains of may not have been seen by the person who is called the writer
of the article until after the appearance of the journal; nay, if he be one
of those--few, I daresay--who do not read their own articles, may never
have been seen by him at all. Possibly, the insertion or omission would not
have been made if the editor could have had one minute's conversation with
his contributor. Sometimes it actually contradicts something which is {18}
allowed to remain in another part of the article; and sometimes, especially
in the case of omission, it renders other parts of the article
unintelligible. These are disadvantages of the system, and a judicious
editor is not very free with his _unus et alter pannus_. Next, readers in
general, when they see the pages of a journal with the articles