I drew on a square card a circle of about 2-1/2 inches in diameter,
and two cross lines through its centre, cutting one another at right
angles. Round each of the four points, 90° apart, where the cross
cuts the circle, I drew small circles of the size of wafers and
gummed upon each a disc of different tint. Finally I made a single
black dot half-way between two of the arms of the cross. I then made
a composite of the four positions of the card, as it was placed
successively with each of its sides downwards. The result is a
photograph having a sharply-defined cross surrounded by four discs
of precisely uniform tint, and between each pair of arms of the
cross there is a very faint dot. This photograph shows many things.
The fact of its being a composite is shown by the four faint dots.
The equality of the successive periods of exposure is shown by the
equal tint of the four dots. The accuracy of adjustment is shown by
the sharpness of the cross being as great in the composite as in the
original card. We see the smallness of the effect produced by any
trait, such as the dot, when it appears in the same place in only
one of the components: if this effect be so small in a series of
only four components, it would certainly be imperceptible in a much
larger series. Thirdly, the uniformity of resulting tint in the
composite wafer is quite irrespective of the order of exposure. Let
us call the four component wafers A, B, C, D, respectively, and the
four composite wafers 1, 2, 3, 4; then we see, by the diagram, that
the order of exposure has differed in each case, yet the result is
identical. Therefore the order of exposure has no effect on the
result.
|----------+------------------------------------|
|Composite.|Successive places of the Components.|
| 1 2 | A B | D A | C D | B C |
| 4 3 | D C | C B | B A | A D |
|===============================================|
In 1 it has been A, D, C, B,
" 2 " B, A, D, C,
" 3 " C, B, A, D,
" 4 " D, C, B, A,
I will next show a series consisting of two portraits considerably
unlike to one another, and yet not so very discordant as to refuse
to conform, and of two intermediate composites. In making one of the
composites I gave two-thirds of the total time of exposure to the
first portrait, and one-third to the second portrait. In making the
other composite, I did the converse. It will be seen how good is the
result in both cases, and how the likeness of the longest exposed
portrait always predominates.
The next is a series of four composites. The first consists of 57
hospital patients suffering under one or other of the many forms of
consumption. I may say that, with the aid of Dr. Mahomed, I am
endeavouring to utilise this process to elicit the physiognomy of
disease. The composite I now show is what I call a hotch-pot
composite; its use is to form a standard whence deviations towards
any particular sub-type may be conveniently gauged. It will be
observed that the face is strongly marked, and that it is quite
idealised. I claim for composite portraiture, that it affords a
method of obtaining _pictorial averages_, which effects
simultaneously for every point in a picture what a method of numerical
averages would do for each point in the picture separately. It
gives, in short, the average tint of every unit of area in the
picture, measured from the fiducial lines as co-ordinates. Now every
statistician knows, by experience, that numerical averages usually
begin to agree pretty fairly when we deal with even twenty or thirty
cases. Therefore we should expect to find that any groups of twenty
or thirty men of the same class would yield composites bearing a
considerable likeness to one another. In proof that this is the case,
I exhibit three other composites: the one is made from the first 28
portraits of the 57, the second from the last 27, and the third is
made from 36 portraits taken indiscriminately out of the 57. It will
be observed that all the four composites are closely alike.
I will now show a few typical portraits I selected out of 82 male
portraits of a different series of consumptive male patients; they
were those that had more or less of a particular wan look, that I
wished to elicit. The selected cases were about 18 in number, and
from these I took 12, rejecting about six as having some marked
peculiarity that did not conform well with the remaining 12. The
result is a very striking face, thoroughly ideal and artistic, and
singularly beautiful. It is, indeed, most notable how beautiful all
composites are. Individual peculiarities are all irregularities, and
the composite is always regular.
I show a composite of 15 female faces, also of consumptive patients,
that gives somewhat the same aspect of the disease; also two others
of only 6 in each, that have in consequence less of an ideal look,
but which are still typical. I have here several other typical faces
in my collection of composites; they are all serviceable as
illustrations of this memoir, but, medically speaking, they are only
provisional results.
I am indebted to Lieutenant Leonard Darwin, R.E., for an interesting
series of negatives of officers and privates of the Royal Engineers.
Here is a composite of 12 officers; here is one of 30 privates. I
then thought it better to select from the latter the men that came
from the southern counties, and to again make a further selection of
11 from these, on the principle already explained. Here is the
result. It is very interesting to note the stamp of culture and
refinement on the composite officer, and the honest and vigorous but
more homely features of the privates. The combination of these two,
officers and privates together, gives a very effective physiognomy.
Let it be borne in mind that existing cartes-de-visite are almost
certain to be useless. Among dozens of them it is hard to find three
that fulfil the conditions of similarity of aspect and of shade. The
negatives have to be made on purpose. I use a repeating back and a
quarter plate, and get two good-sized heads on each plate, and of a
scale that never gives less than four-tenths of an inch between the
pupils of the eyes and the mouth. It is only the head that can be
used, as more distant parts, even the ears, become blurred hopelessly.
It will be asked, of what use can all this be to ordinary
photographers, even granting that it may be of scientific value in
ethnological research, in inquiries into the physiognomy of disease,
and for other special purposes? I think it can be turned to most
interesting account in the production of family likenesses. The most
unartistic productions of amateur photography do quite as well for
making composites as those of the best professional workers, because
their blemishes vanish in the blended result. All that amateurs have
to do is to take negatives of the various members of their families
in precisely the same aspect (I recommend either perfect full-face
or perfect profile), and under precisely the same conditions of
light and shade, and to send them to a firm provided with proper
instrumental appliances to make composites from them. The result is
sure to be artistic in expression and flatteringly handsome, and
would be very interesting to the members of the family. Young and old,
and persons of both sexes can be combined into one ideal face. I can
well imagine a fashion setting in to have these pictures.
Professional skill might be exercised very effectively in retouching
composites. It would be easy to obliterate the ghosts of stray
features that are always present when the composite is made from
only a few portraits, and it would not be difficult to tone down any
irregularity in the features themselves, due to some obtrusive
peculiarity in one of the components. A higher order of artistic
skill might be well bestowed upon the composites that have been made
out of a large number of components. Here the irregularities
disappear, the features are perfectly regular and idealised, but the
result is dim. It is like a pencil drawing, where many attempts have
been made to obtain the desired effect; such a drawing is smudged
and ineffective; but the artist, under its guidance, draws his final
work with clear bold touches, and then he rubs out the smudge. On
precisely the same principle the faint but beautifully idealised
features of these composites are, I believe, capable of forming the
basis of a very high order of artistic work.
B.--THE RELATIVE SUPPLIES FROM TOWN AND COUNTRY FAMILIES
TO THE POPULATION OF FUTURE GENERATIONS.
[_Read before the Statistical Society in_ 1873.]
It is well known that the population of towns decays, and has to be
recruited by immigrants from the country, but I am not aware that
any statistical investigation has yet been attempted of the rate of
its decay. The more energetic members of our race, whose breed is
the most valuable to our nation, are attracted from the country to
our towns. If residence in towns seriously interferes with the
maintenance of their stock, we should expect the breed of Englishmen
to steadily deteriorate, so far as that particular influence is
concerned.
I am well aware that the only perfectly trustworthy way of
conducting the inquiry is by statistics derived from numerous
life-histories, but I find it very difficult to procure these data.
I therefore have had recourse to an indirect method, based on a
selection from the returns made at the census of 1871, which appears
calculated to give a fair approximation to the truth. My object is
to find the number of adult male representatives in this generation,
of 1000 adult males in the previous one, of rural and urban
populations respectively. The principle on which I have proceeded is
this:--
I find (A) the number of children of equal numbers of urban and of
rural mothers. The census schedules contain returns of the names and
ages of the members of each "family," by which word we are to
understand those members who are alive and resident in the same
house with their parents. When the mothers are young, the children
are necessarily very young, and nearly always (in at least those
classes who are unable to send their children to boarding schools)
live at home. If, therefore, we limit our inquiries to the census
"families" of young mothers, the results may be accepted as
practically identical with those we should have obtained if we had
direct means of ascertaining the number of their living children.
The limits of age of the mothers which I adopted in my selection were,
24 and 40 years. Had I to begin the work afresh, I should prefer
the period from 20 to 35, but I have reason to feel pretty well
contented with my present data. I correct the results thus far
obtained on the following grounds:--(B) the relative mortality of
the two classes between childhood and maturity; (C) the relative
mortality of the rural and urban mothers during childbearing ages;
(D) their relative celibacy; and (E) the span of a rural and urban
generation. It will be shown that B is important, and C noteworthy,
but that D and E may be disregarded.
In deciding on the districts to be investigated, it was important to
choose well-marked specimens of urban and rural populations. In the
former, a town was wanted where there were various industries, and
where the population was not increasing. A town where only one
industry was pursued would not be a fair sample, because the
particular industry might be suspected of having a special influence,
and a town that was increasing would have attracted numerous
immigrants from the country, who are undistinguishable as such in
the census returns. Guided by these considerations, I selected
Coventry, where silk weaving, watch-making, and other industries are
carried on, and whose population had scarcely varied during the
decade preceding the census of 1871.[25] It is an open town, in
which the crowded alleys of larger places are not frequent. Its urban
peculiarities are therefore minimised, and its statistical returns
would give a picture somewhat too favourable of the average
condition of life in towns. For specimens of rural districts, I
chose small agricultural parishes in Warwickshire.
[Footnote 25: It has greatly changed since this was written.]
By the courteous permission of Dr. Farr, I was enabled to procure
extracts from the census returns concerning 1000 "families" of
factory hands at Coventry, in which the age of the mother was
neither less than 24 nor more than 40 years, and concerning another
1000 families of agricultural labourers in rural parishes of
Warwickshire, under the same limitations as to the age of the mother.
When these returns were classified (see Table I., p. 246), I found
the figures to run in such regular sequence as to make it certain
that the cases were sufficiently numerous to give trustworthy results.
It appeared that:
(A) The 1000 families of factory hands comprised 2681 children, and
the 1000 of agricultural labourers comprised 2911; hence, the
children in the urban "families," the mothers being between the ages
of 24 and 40, are on the whole about 8 per cent, less numerous than
the rural. I see no reason why these numbers should not be accepted
as relatively correct for families, in the ordinary sense of that
word, and for mothers of all ages. An inspection of the table does
indeed show that if the selection had begun at an earlier age than 24,
there would have been an increased proportion of sterile and of
small families among the factory hands, but not sufficient to
introduce any substantial modification of the above results. It is,
however, important to recollect that the small error, whatever its
amount may be, is a concession in favour of the towns.
(B) I next make an allowance for the mortality between childhood and
maturity, which will diminish the above figures in different
proportions, because the conditions of town life are more fatal to
children than those of the country. No life tables exist for
Coventry and Warwickshire; I am therefore obliged to use statistics
for similarly conditioned localities, to determine the amount of the
allowance that should be made. The life tables of Manchester [26]
will afford the data for towns, and those of the "Healthy Districts"
[27] will suffice for the country. By applying these, we could
calculate the number of the children of ages specified in the census
returns who would attain maturity. I regret extremely that when I
had the copies taken, I did not give instructions to have the ages
of all the children inserted; but I did not, and it is too late now
to remedy the omission. I am therefore obliged to make a very rough,
but not unfair, estimate. The average age of the children was about
3 years, and 25 years may be taken as representing the age of
maturity. Now it will be found that 74 per cent. of children in
Manchester, of the age of 3, reach the age of 25, while 86 per cent.
of children do so in the "Healthy Districts." Therefore, if my rough
method be accepted as approximately fair, the number of adults who
will be derived from the children of the 1000 factory families
should be reckoned at (2681 × 74/100) = 1986, and those from the
1000 agricultural at (2911 × 86/100) = 2503.
[Footnote 26: "Seventh Annual Report of Registrar-General."]
[Footnote 27: Healthy Districts Life Table, by Dr. Farr. _Phil
Trans. Royal Society_, 1859.]
(C) The comparison we seek is between the total families produced by
an equal number of urban and rural women who had survived the age of
24. Many of these women will not marry at all; I postpone that
consideration to the next paragraph. Many of the rest will die
before they reach the age of 40, and more of them will die in the
town than in the country. It appears from data furnished by the
above-mentioned tables, that if 100 women of the age of 24 had
annually been added to a population, the number of those so added,
living between the ages of 24 and 40 (an interval of seventeen years)
would be 1539 under the conditions of life in Manchester, and 1585
under those of the healthy districts. Therefore the small factors to
be applied respectively to the two cases, on account of this
correction, are 1539/(17 × 100) and 1585/(17 × 100).
(D) I have no trustworthy data for the relative prevalence of
celibacy in town and country. All that I have learned from the
census returns is, that when searching them for the 1000 families,
131 bachelors were noted between the ages of 24 and 40, among the
factory hands, and 144 among the agricultural labourers. If these
figures be accepted as correct guides to the amount of celibacy
among the women, it would follow that I must be considered to have
discussed the cases of 1131 factory, and 1144 agricultural women,
when dealing with those of 1000 mothers in either class.
Consequently that the respective corrections to be applied, are
given by the factors 1000/1131 and 1000/1141 or 88.4/1000 and 87.6/
1000. This difference of less than 1 per cent, is hardly worth
applying, moreover I do not like to apply it, because it seems to me
erroneous and to act in the wrong direction, inasmuch as unmarried
women can obtain employment more readily in the town than in the
country, and celibacy is therefore more likely to be common in the
former than in the latter.
(E) The possible difference in the length of an urban and rural
generation must not be forgotten. We, however, have reason to
believe that the correction on this ground will be insignificant,
because the length of a generation is found to be constant under
very different circumstances of race, and therefore we should expect
it to be equally constant in the same race under different conditions;
such as it is, it would probably tell against the towns.
Let us now sum up the results. The corrections are not to be applied
for (D) and (E), so we have only to regard (A) × (B) × (C), that
this--
2681 × 74/100 × 1539/1700 1796 77
------------------------- = ---- = --
2911 × 86/100 × 1585/1700 2334 100
In other words, the rate of supply in towns to the next adult
generation is only 77 per cent., or, say, three-quarters of that in
the country. This decay, if it continued constant, would lead to the
result that the representatives of the townsmen would be less than
half as numerous as those of the country folk after one century, and
only about one fifth as numerous after two centuries, the
proportions being 45/100 and 21/100 respectively.
[Transcriber's Note: In the original manuscript, Table I occupied
two facing pages. This is the left-hand (sinister) page; the right-hand
(dexter) page is immediately below.]
TABLE I. -- _Census Returns of 1000 Families of Factory Hands in
Coventry, and 1000 Families of Agricultural Labourers in Warwickshire,
grouped according to the Age of the Mother and the Number of Children
in the Family._
---------------------------------------------------
|NUMBER OF CHILDREN IN FAMILY. |
|---------|---------|---------|----------|--------|
| 0. | 1. | 2. | 3. | 4. |
|---------+---------+---------+----------+--------|
| F | A | F | A | F | A | F | A | F | A |
| a | g | a | g | a | g | a | g | a | g |
| c | r | c | r | c | r | c | r | c | r |
| t | i | t | i | t | i | t | i | t | i |
| o | c | o | c | o | c | o | c | o | c |
| r | u | r | u | r | u | r | u | r | u |
Age of Mother | y | l | y | l | y | l | y | l | y | l |
| . | t | . | t | . | t | . | t | . | t |
---------------------------------------------------
24 to 25 | 28 17 40 31 | 24 32 12 10 2 |
| +-------------------+ |
26 " 27 | 19 18 36 24 36 28 23 26 | 8 8 |
| | |
28 " 29 | 18 17 32 16 20[A] 33 36 23 | 14 23 |
| | |
30 " 31 | 13 4 23 18 24 21 28[A] 31 | 18 22 |
| | |
32 " 33 | 18 11 16 14 19 13 22[A] 27 | 23 26 |
|---------+ | |
34 " 35 | 14 15 | 11 6 17 16 28 18 | 31 34 |
| +-------------------+ | |
36 " 37 | 12 17 4 11 10 13 | 22 14 | 16 20 |
| +---------+ |
38 " 39 | 8 6 9 15 14 17 16 21 22 23 |
| |
40 | 8 7 3 10 8 9 13 14 8 10 |
===============|=================================================|
Total within | |
outline | 96 67 258 109 116 111 171 149 |
Total between | |
outlines | 42 45 16 36 56 71 29 35 142 166 |
Total beyond | |
outline | |
===============|=================================================|
Total |138 112 174 145 172 182 200 184 142 166 |
===============|=================================================|
[Footnote A: These three cases are anomalous, the Factory being less
than the Agricultural. In the instance of 20-33, the anomaly is double,
because the sequence of the figures shows that neither of these can be
correct; certainly not the first of them.]
_Note_.--It will be observed to the left of the outline, that is,
in the upper and left hand of the table, where the mothers are young
and the children few, the factory families predominate, while the
agricultural are the most numerous between the outlines, that is,
especially in the middle of the table, where the mothers are less young,
and the family is from four to five in number. The two are equally
numerous to the right of the outlines, that is, to the right of the
table, where the families are large.
[Transcriber's Note: In the original manuscript, Table I occupied
two facing pages. This is the right-hand (dexter) page; the left-hand
(snister) page is immediately above.]
TABLE I. -- _Census Returns of 1000 Families of Factory Hands in
Coventry, and 1000 Families of Agricultural Labourers in Warwickshire,
grouped according to the Age of the Mother and the Number of Children
in the Family._
| NUMBER OF CHILDREN IN FAMILY. |
|-------------------------------------------------|
| 5. | 6. | 7. | 8. | 9. |
|---------+---------+---------+---------+---------|
| F | A | F | A | F | A | F | A | F | A |
| a | g | a | g | a | g | a | g | a | g |
| c | r | c | r | c | r | c | r | c | r |
| t | i | t | i | t | i | t | i | t | i |
| o | c | o | c | o | c | o | c | o | c |
| r | u | r | u | r | u | r | u | r | u |
| y | l | y | l | y | l | y | l | y | l |Age of Mother
| . | t | . | t | . | t | . | t | . | t |
|---------+---------+---------+---------+---------|------------
| 1 1 | | 24 to 25
| | |
| | | 26 " 27
| | |
| 6 6 | 4 1 2 | 28 " 29
| | |
| 12 15 | 2 5 2 1 | 30 " 31
| | |
| 21 25 | 9 5 1 2 | 32 " 33
| | |
| 14 18 | 12 9 5 3 1 | 34 " 35
| | |
| 15 25 | 12 10 4 5 5 2 | 36 " 37
| | |
| 14 22 | 10 15 6 7 2 1 | 38 " 39
| | |
| 7 11 | 3 9 7 7 2 1 | 40
|=================================================|--------------------
| |Total within outline.
| 90 123 |Total between outline
| 52 54 24 25 7 9 1 |Total beyond outline.
|=================================================|=====================
| 90 123 52 54 24 25 7 9 1 |Total.
|=======================================================================
TABLE II.
|----------------------------------------------------------------------|
| | Number of Families | Number of Children |
| |--------+--------------+------------------------|
| | Factory| Agricultural | Factory | Agricultural |
| Within outline | 541 | 436 | 903 | 778 |
| Between outlines | 375 | 476 | 1233 | 1562 |
| Beyond outlines | 84 | 88 | 545 | 571 |
|=============================================+========================|
| Total | 1000 | 1000 | 2681 | 2911 |
|======================================================================|
C -- AN APPARATUS FOR TESTING THE DELICACY WITH WHICH WEIGHTS CAN BE
DISCRIMINATED BY HANDLING THEM.
[_Read at the Anthropological Institute_, Nov., 1882.]
I submit a simple apparatus that I have designed to measure the
delicacy of the sensitivity of different persons, as shown by their
skill in discriminating weights, identical in size, form, and colour,
but different in specific gravity. Its interest lies in the
accordance of the successive test values with the successive
graduations of a true scale of sensitivity, in the ease with which
the tests are applied, and the fact that the same principle can be
made use of in testing the delicacy of smell and taste.
I use test-weights that mount in a series of "just perceptible
differences" to an imaginary person of extreme delicacy of perception,
their values being calculated according to Weber's law. The lowest
weight is heavy enough to give a decided sense of weight to the hand
when handling it, and the heaviest weight can be handled without any
sense of fatigue. They therefore conform with close approximation to
a geometric series; thus--
_WR0, WR1, WR2, WR3_, etc.,
and they bear as register-marks the values of the successive indices,
0, 1, 2, 3, etc. It follows that if a person can just distinguish
between any particular pair of weights, he can also just distinguish
between any other pair of weights whose register-marks differ by the
same amount. Example: suppose A can just distinguish between the
weights bearing the register-marks 2 and 4, then it follows from the
construction of the apparatus that he can just distinguish between
those bearing the register-marks 1 and 3, or 3 and 5, or 4 and 6, etc.;
the difference being 2 in each case.