. R. Stuart Poole, the learned
curator of the coins and gems in the British Museum, for his kind
selection of the most suitable medals, and for procuring casts of
them for me for the present purpose. These casts were, with one
exception, all photographed to a uniform size of four-tenths of an
inch between the pupils of the eyes and the division between the lips,
which experience shows to be the most convenient size on the whole
to work with, regard being paid to many considerations not worth
while to specify in detail. When it was necessary the photograph was
reversed. These photographs were made by Mr. H. Reynolds; I then
adjusted and prepared them for taking the photographic composite.

The next series to be exhibited consists of composites taken from
the portraits of criminals convicted of murder, manslaughter, or
crimes accompanied by violence. There is much interest in the fact
that two types of features are found much more frequently among
these than among the population at large. In one, the features are
broad and massive, like those of Henry VIII., but with a much
smaller brain. The other, of which five composites are exhibited,
each deduced from a number of different individuals, varying four to
nine, is a face that is weak and certainly not a common English face.
Three of these composites, though taken from entirely different sets
of individuals, are as alike as brothers, and it is found on
optically combining any three out of the five composites, that is on
combining almost any considerable number of the individuals, the
result is closely the same. The combination of the three composites
just alluded to will now be effected by means of the three
converging magic-lanterns, and the result may be accepted as generic
in respect of this particular type of criminals.

The process of composite portraiture is one of pictorial statistics.
It is a familiar fact that the average height of even a dozen men of
the same race, taken at hazard, varies so little, that for ordinary
statistical purposes it may be considered constant. The same may be
said of the measurement of every separate feature and limb, and of
every tint, whether of skin, hair, or eyes. Consequently a pictorial
combination of any one of these separate traits would lead to
results no less constant than the statistical averages. In a portrait,
there is another factor to be considered besides the measurement of
the separate traits, namely, their relative po