iality, unreality, sophistry and catch-pennyism of the one
division with the priggishness, the lack of tact and humour, and above
all the pseudo-scientific tendency to generalisation, classification
and, to use a familiar word, "pottering" of the other. In particular he
had a mania in his more serious moods for defining and sub-defining
things and putting them into pigeon-holes under the sub-definitions.
Thus the so-called Demetrius Phalereus, who (or a false namesake of his)
has left us a capital _general_ remark (to be given presently) on
letter-writing, elaborately divides its kinds, with prescriptions for
writing each, into "friendly," "commendatory," "reproving,"
"objurgatory," "consolatory," "castigatory," "admonishing,"
"threatening," "vituperatory," "laudatory," "persuasive," "begging,"
"questioning," "answering," "allegorical," "explanatory," "accusing,"
"defending," "congratulatory," "ironic" and "thankful," while the
neo-Platonist, Proclus, is responsible for, or at least has attributed
to him, a list of nearly double the length, including most of those
given above and adding many. Of these last, "love-letters" is the most
important, and "mixed" the _canniest_, for it practically lets in
everything.

This way, of course, except for purely business purposes--where
established forms save time, trouble and possible litigation--no
possible good lies; and indeed the impossibility thereof is clearly
enough indicated in the above-glanced-at general remark of Demetrius (or
whoever it was) himself. In fact the principle of this remark and its
context in the work called "Of Interpretation," which it is more usual
now to call, perhaps a little rashly, "Of Style," is so different from
the catalogue of types that they can hardly come from the same author.
"You _can_ from this, as well as from all other kinds of writing,
discern the character of the writer; indeed from none other can you
discern it so well." Those who know a little of the history of Criticism
will see how this anticipates the most famous and best definitions of
Style itself, as being "the very man," and they may perhaps also think
worthy of notice another passage in the same context where the author
finds fault with a rather "fine" piece of an epistle as "not the way a
man would talk to his friend," and even goes on to use the most familiar
Greek word for talking--[Greek: lalein]--in the same connection.

[Sidenote: ALCIPHRON. JULIAN]

Of such "talking