towards what one may call a severe and ascetic view
of history. While writing _The Making of England_ and _The Conquest
of England_, he used to lament the scantiness of the data and the
barren dryness which he feared the books would consequently show.
"How am I to make anything of these meagre entries of marches and
battles which are the only materials for the history of whole
centuries? Here are the Norsemen and Danes ravaging and occupying
the country; we learn hardly anything about them from English
sources, and nothing at all from Danish. How can one conceive and
describe them? how have any comprehension of what England was like in
the districts the Northmen took and ruled?" I tried to get him to
work at the Norse Sagas, and remember in particular to have
entreated him when he came to the battle of Brunanburh to eke out
the pitifully scanty records of that fight from the account given of
it in the story of the Icelandic hero, Egil, son of Skallagrim. But
he answered that the Saga was unhistorical, a bit of legend
written down more than a century after the events, and that he
could not, by using it in the text, appear to trust it, or to mix up
authentic history with what was possibly fable. It was urged that he
could guard himself in a note from being supposed to take it for more
than what it was, a most picturesque embellishment of his tale. But
he stood firm. Throughout these two last books, he steadily
refrained from introducing any matter, however lively or romantic,
which could not stand the test of his stringent criticism, and used
laughingly to tell how Dean Stanley had long ago said to him, after
reading one of his earliest pieces, "I see you are in danger of
growing picturesque. Beware of it. I have suffered for it."

If in these later years he reined in his imagination more tightly, the
change was due to no failing in his ingenuity. Nothing in his work
shows higher constructive ability than _The Making of England_. He had
to deal with a time which has left us scarcely any authentic records,
and to piece together his narrative and his picture of the country out
of these records, and the indications, faint and scattered, and often
capable of several interpretations, which are supplied by the remains
of Roman roads and villas, the names of places, the boundaries of
local divisions, the casual statements of writers many centuries
later. What he has given us remains an enduring witness to his
historical power.