details; he must, in the midst of such occupations, preserve unity of
design, as much as must the novelist or the playwright; and yet with all
this there is not a verb, an adjective or a substantive which, if it
does not repose upon established evidence, will not mar the particular
type of work on which he is engaged.

As an example of what I mean, consider two sentences: The first is taken
from the 432nd page of that exceedingly unequal publication, the
_Cambridge History of the French Revolution_; the second I have
made up on the spur of the moment; both deal with the Battle of
Wattignies. The "Cambridge History" version runs as follows:--

On October 15 the relieving force, 50,000 strong, attacked the Austrian
covering force at Wattignies; the battle raged all that day and was most
furious on the right, in front of the village of Wattignies, which was
taken and lost three times; on the 17th the French expected another
general engagement but the enemy had drawn off.

There are here five great positive errors in six lines. The French were
not 50,000 strong, the attack on the 15th was not on Wattignies, but on
Dourlers; Wattignies was not taken and lost three times; the fight of
the 15th was _least_ pressed on the right (harder on the left and
hardest in the centre) and no one--not the least recruit--expected
Coburg to come _back_ on the 17th. Why, he had crossed the Sambre
at every point the day before! As for negative errors, or errors of
omission, they are capital, and the chief is that the victory was won on
the second day, the 16th, of which no mention is made.

Now contrast such a sentence with the following:--