s, such as I have just now proposed. For this analysis you will
take some passage of English verse or prose--say the first ten lines of
_Paradise Lost_--or the Lord's Prayer--or the 23rd Psalm; you will
distribute the whole body of words contained in that passage, of course
not omitting the smallest, according to their nationalities--writing, it
may be, A over every Anglo-Saxon word, L over every Latin, and so on
with the others, if any other should occur in the portion which you have
submitted to this examination. When this is done, you will count up the
_number_ of those which each language contributes; again, you will note
the _character_ of the words derived from each quarter.

{Sidenote: _Two Shapes of Words_}

Yet here, before I pass further, I would observe in respect of those
which come from the Latin, that it will be desirable further to mark
whether they are directly from it, and such might be marked L1, or only
mediately from it, and to us directly from the French, which would be
L2, or L at second hand--our English word being only in the second
generation descended from the Latin, not the child, but the child's
child. There is a rule that holds pretty constantly good, by which you
may determine this point. It is this,--that if a word be directly from
the Latin, it will not have undergone any alteration or modification in
its form and shape, save only in the termination--'innocentia' will
have become 'innocency', 'natio' will have become 'nation',
'firmamentum' 'firmament', but nothing more. On the other hand, if it
comes _through_ the French, it will generally be considerably altered in
its passage. It will have undergone a process of lubrication; its
sharply defined Latin outline will in good part have departed from it;
thus 'crown' is from 'corona', but though 'couronne', and itself a
dissyllable, 'coroune', in our earlier English; 'treasure' is from
'thesaurus', but through 'tresor'; 'emperor' is the Latin 'imperator',
but it was first 'empereur'. It will often happen that the substantive
has past through this process, having reached us through the
intervention of the French; while we have only felt at a later period
our want of the adjective also, which we have proceeded to borrow direct
from the Latin. Thus, 'people' is indeed 'populus', but it was 'peuple'
first, while 'popular' is a direct transfer of a Latin vocable into our
English glossary. So too 'enemy' is 'inimicus', but it was first
softened i