such MSS.
as name the 18th of April ought to be respected; but Tyrwhitt's "28th,"
which he states not only as the result of his own conjecture but as
authorised by the "the best MSS.," ought to be scouted at once.

In the latest edition of the "Canterbury Tales" (a literal reprint from one
of the Harl. MSS., for the Percy Society, under the supervision of Mr.
Wright), the opening of the Prologue to "The Man of Lawes Tale" does not
materially differ from Tyrwhitt's text, excepting in properly assigning the
day of the journey to "the eightetene day of April;" and the confirmation
of the forenoon altitude is as follows:

"And sawe wel that the schade of every tree
Was in the lengthe the same quantite,
That was the body erecte that caused it."

But the afternoon observation is thus related:

"By that the Manciple had his tale endid,
The sonne fro the southe line is descendid
So lowe that it nas nought to my sight,
Degrees nyne and twenty as in hight.
_Ten_ on the clokke it was as I gesse,
For eleven foote, or litil more or lesse,
My schadow was at thilk time of the yere,
Of which feet as my lengthe parted were,
In sixe feet equal of proporcioun."

In a note to the line "Ten on the clokke" Mr. Wright observes,

"_Ten_. I have not ventured to change the reading of the Harl. MS.,
which is partly supported by that of the lands. MS., _than_."

If the sole object were to present an exact counterpart of the MS., of
course even its errors were to be respected: but upon no other grounds can
I understand why a reading should be preserved by which broad sunshine is
attributed to ten o'clock at night! Nor can I believe that the copyist of
the MS. with whom the error must have originated would have set down
anything so glaringly absurd, unless he had in his own mind some means of
reconciling it with probability. It may, I believe, be explained in the
circumstance that "ten" and "four," in horary reckoning, were _convertible
terms_. The old Roman method of naming the hours, wherein noon was the
sixth, was long preserved, especially in conventual establishments: and I
have no doubt that the English idiomatic phrase "o'clock" originated in the
necessity for some distinguishing mark between hours "of the clock"
reckoned from midnight, and hours of the day reckoned from sunrise, or more
frequently from six A.M. With such an understanding, it is clear that _ten_
might be called _four_, and _four ten_,