must be coloured and shaded with many touches before it can really
resemble the face, and whilst this is being done the maintenance of
the likeness is imperilled at every step. I lately watched an able
artist painting a portrait, and endeavoured to estimate the number
of strokes with his brush, every one of which was thoughtfully and
firmly given. During fifteen sittings of three working hours
each--that is to say, during forty-five hours, or two thousand four
hundred minutes--he worked at the average rate of ten strokes of the
brush per minute. There were, therefore, twenty-four thousand
separate traits in the completed portrait, and in his opinion some,
I do not say equal, but comparably large number of units of
resemblance with the original.

The physiognomical difference between different men being so
numerous and small, it is impossible to measure and compare them
each to each, and to discover by ordinary statistical methods the
true physiognomy of a race. The usual way is to select individuals
who are judged to be representatives of the prevalent type, and to
photograph them; but this method is not trustworthy, because the
judgment itself is fallacious. It is swayed by exceptional and
grotesque features more than by ordinary ones, and the portraits
supposed to be typical are likely to be caricatures. One fine Sunday
afternoon I sat with a friend by the walk in Kensington Gardens that
leads to the bridge, and which on such occasions is thronged by
promenaders. It was agreed between us that whichever first caught
sight of a typical John Bull should call the attention of the other.
We sat and watched keenly for many minutes, but neither of us found
occasion to utter a word.

The prevalent type of English face has greatly changed at different