PRESERVATION OF OBJECTS
The preservation of the objects that are found is a necessary duty
of the finder. To disclose things only to destroy them, when a more
skilful or patient worker might have added them to the world’s
treasures, is a hideous fault. And the excavator must be ready for all
emergencies, for all classes of objects in all stages of decay, and
deal with each without delays, and often with scanty and unsuitable
means at hand for their treatment. Some familiarity with chemistry and
physics and properties of materials, is one of the first requisites for
an excavator. All this applies in a lesser degree to the difficulties
of transport, which is also part of the preservation of the antiquities.
As conditions so infinitely vary it is useless to lay down any fixed
rules for treatment. Such rules would hinder the use of common sense,
which is essential to success. But examples of how different materials
are affected, and how difficulties have been met, will lead to the
excavator thinking out a fit treatment for each case as it arises. In
all this we are stating field practice only, and not dealing with
museum methods, which differ by having far more command of resources,
and by not having to deal with any of the troublesome cases which do
not survive to reach a museum.
[Sidenote: Stone.]
The great enemy of stonework is salt. In Egypt this permeates the
soil so that nothing is free from it; and any object near the surface
has much salt accumulated in it by evaporation. The effect of salt is
to disintegrate the stone, and make it flake or fall away in powder.
If there is the faintest taste of salt on a stone slab it should be
laid to dry, face down, on the ground; for I have seen a fine block
of sculpture entirely destroyed by being left for a single day face
upward. When the stone is once dry it is safe in Egypt, but in a
damp country it may begin a course of slow destruction by continual
recrystallization of salt. Sculptures have been entirely wrecked by
being cemented into the wall of a museum; the wet of the cement brought
all the salt to the face and ruined it. The only treatment for salt in
stone or any other material is long soaking in water. If a canal is at
hand, stones may be sunk in it for some weeks, face down. Or barrels
or zinc trays may be used, and the water changed every two or three
days, for five or six times. After such soaking the stone must be left
to dry face down, so that all the remaining salt will come out on the
back. Where there is not much salt it would be best to lay the stone
back upwards to dry, brush off any salt which comes out, and then wet
the ground below, so that more water may be drawn up to evaporate on
the back. If this was continued until no salt appeared the stone would
be cleaned, and the face could not be injured. Sometimes a face is
already flaking, and then the stone must be kept quite flat in soaking
and drying, so that each flake will be left in place, and can be stuck
down afterwards. Granite is often entirely disintegrated into separate
crystals, if it has lain near the surface. It is then even impossible
to turn the block over to copy it, as there is no cohesion left in the
mass. The only salvation possible for such a block would be to make a
thick plaster or cement coat to the exposed parts, under cut, and turn
the whole over with a board beneath it, and then saturate it hot with
paraffin wax.
The face of limestone is often in tender condition, and will not bear
wet brushing to clean it. Dry picking and brushing is then the only
resource. If long exposed to damp, limestone dissolves throughout
the body of it, so that it becomes spongy, and like putty with the
contained water. A large sarcophagus lid in this state at Denderah was
brought up to the house, then covered with 3 or 4 inches of sand, and
left to dry slowly for some weeks; otherwise it would have cracked to
chips by contraction on the face. When quite dry it was very porous,
but in safe state for copying and transport. I have seen a slab of
limestone in perfect condition, reduced to a shapeless paste by a few
minutes of sharp rain.
The original stucco facing often remains on limestone, and also the
colour. If the carving has been fine it is best to remove the stucco,
which is generally much less detailed. But if the stucco is an
improvement on the carving, and especially if there is colour, it must
be preserved. This is best done by fixing it with thin tapioca water,
just so thick that it will soak into the stone without leaving any
glair when dry. This treatment also does for limestone with a rotten
face.
The same tapioca water may be used for fixing colours on stucco, as I
did on the Tell el Amarna pavement (Fig. 48); and the thickness must be
graduated to the porosity, so that it will just soak entirely into the
material. Any film left on the face will peel away.
[Illustration: FRESCOES, TELL EL AMARNA.
Fig. 48. Plants and animals.]
[Illustration: FRESCOES, TELL EL AMARNA.
Fig. 49. The two princesses.]
[Sidenote: Pottery.]
Pottery has not much to fear except salt, and that should be soaked
out as from stone. Glazed pottery with salt in it is more difficult
to clear, as it takes so long to get any change in and out of it. But
a persistent soaking will clear it in the course of some weeks; and,
if necessary, partly drying it in intervals, will bring the salt out
of the cracks, whence it can be dusted off. The commonest failing of
glazes is decomposition. The green turn brown, by the decomposition
of the iron from green silicate to brown oxide; and this may take
place from the porous interior without breaking the external face. The
blue glazes go white; and this can be partly remedied by warming and
soaking with paraffin wax, which fills the fine cracks and displays the
remaining colour again. Sometimes the outer coat of clear glaze over
faience inlay is decomposed, without spoiling the faience below. In
this case it is like a picture of which the varnish is gone brown,--it
only needs cleaning. The decomposed glaze can be scraped off, or rubbed
with fine emery paper, until the faience is clean, and then a coat
of paraffin wax clears the colour and preserves it from decomposition.
When glazed ware, especially of the earliest times, is first found, it
is very tender and soft. It then needs the most careful handling, and
must not be brushed or cleaned until it is quite dry and hardened.
[Sidenote: Textiles.]
Textiles are also often saturated with salt, especially the Coptic
garments which are in graves near the surface. They may be safely
soaked to remove the salt and the organic matter, and then dried by
pressing in a towel and laying between sheets of paper. The most tender
examples might perhaps be best treated by placing with half a dozen
sheets of blotting-paper over and under, and keeping wet below while
evaporating on the top; this would carry the salt out to the top of the
blotting paper. In any long soaking of organic stuffs a little carbolic
acid is desirable, to prevent souring and putrefaction of the material.
In every case the threads of textiles are liable to crumble, and any
great amount of washing will tend to reduce a good deal to powder.
Ironing is always desirable to consolidate the stuff.
[Sidenote: Wood.]
Wood does not suffer so much from salt as from rot and white ants. Any
salt may be soaked out; or, if the wood is tender and will not bear
that, a very stiff jelly should be made, so that it will just melt at
boiling: the wood dropped in when the jelly liquefies, and left in the
jelly cold for a week or two. Then the salt will dialyse out into the
jelly, without any free water softening the wood. On remelting the
jelly the wood can be removed, and the salt will be left in the jelly.
The gelatine will strengthen and improve the wood. This process can be
used excellently for ivories or bones, which would be ruined by soaking
in water. Whole skeletons can be set in stiff size, and taken out weeks
after, freed from salt, as was done to those from Medum, now in the
College of Surgeons.
Rotted wood is very tender to handle; and from its continued
contraction when exposed to the air it will fall to pieces. If nearly
dry, but rotted, the best safeguard is to coat it with beeswax or
paraffin wax; if it can be lifted threads can be slipped round it,
and the whole dipped in hot wax until soaked. Or it may have a rapid
coat of wax chilled upon it, which protects it and binds it together
for travelling, and which can be soaked into it by piecemeal heating
afterwards. If the wood will not bear lifting, it may be coated by
dashing on superheated paraffin wax almost at boiling-point. This
will soak deep into the wood like hot water, and consolidate it so
that it can be moved quite safely. The same processes apply also to
stuccoed wood, which needs such safeguards, as otherwise the stucco all
falls off by the continued shrinkage of the wood. The great stuccoed
sarcophagus at Hawara was preserved by heating the surface with a wire
dish of charcoal burning about six inches above it, and flooding the
surface with melted wax so soon as it was enough heated to absorb it.
Perhaps superheated paraffin wax would have carried enough heat with it
to soak in without the charcoal fire. For all heating of wax it is best
to use a cast-iron saucepan, as soldered tins may give way before the
wax boils. Another treatment, especially suited for large objects, is
painting with several coats of wax dissolved in benzol.
Wood which is very wet is more difficult to manage. It may be kept for
long under water, like the wood from the Glastonbury lake village. And
it may be consolidated with silicate solution, as has been well done
in examples from Silchester. Or it may be removed from water and laid
in glycerine with the top exposed; thus the water will evaporate and
diffuse, and glycerine take its place.
[Sidenote: Ivory.]
Ivory is mainly liable to flaking, especially if in wet soil. When any
ivory is seen not in a firm condition, the earth should be carefully
worked round so as to find the limits of the ivory, be it a single
piece or a collection together. Then the mass should be under-cut down
to a firm stratum, and lifted out in a whole block of earth. This
should be left to dry slowly; and after a week or two the earth should
be gently brushed away with a camel-hair brush, aided by picking with
a stout pin. As each piece of ivory is seen it should be carefully
followed, and if quite dry it may probably be removed entire. If still
liable to flake, it can then be soaked in melted paraffin wax. If the
ivory is too rotted to be detached from the earth, then the whole
mass would have to be baked to rather over blood heat, and saturated
with paraffin wax. After that it could be safely dissected by careful
picking. In case of finding large groups of ivories in the ground,
too extensive to take out in a block to dry, probably it would do to
isolate them, then lay a few inches of sand on the top, and light a
fire over them: after slow burning for a few days the ground would be
baked dry below, and could be saturated with wax before lifting the
mass.
It sometimes happens that ivories in wet soil get concreted crystalline
carbonate of lime upon them, which is much harder than the ivory.
This being crystalline is not saturated with wax when the ivory is so
treated. Hence after waxing the ivory the surface should be cleaned
with benzol or ether on cotton-wool, and then painted with nitric acid
to dissolve the crystalline lime. Even strong nitric acid will only
dull the surface of waxed ivory, and not remove any perceptible amount,
while it dissolves the concretion rapidly. Probably the darkening of
the ivory caused by soaking in wax can be mainly removed by heating
fuller’s earth to over boiling-point, and then rapidly packing the
ivory in the earth and pressing it: the heat would melt the wax on the
surface, it will be absorbed by the earth, and the face of the ivory
will be left dry of wax. The ivories from Nineveh were solidified with
gelatine; but that would probably break up very tender ivories by the
amount of water. In case however of much salt in ivory the best way to
treat it is to drop it in stiff hot gelatine, cool it, and let it lie
in the consolidated mass for a week or two, for the salt to dialyse
out. Another way, if the mass is not much cut into hollows, is to lash
the ivory closely with thread or fine twine, and then soak it in water
to remove the salt; the twine prevents it falling to pieces, and it can
be dipped in wax when dry, and the twine removed.
[Sidenote: Papyri.]
Papyri require most careful treatment at every stage. They are often
found in a very fragile state, and if the roll has to be carried
without special packing in wool it is best to wrap it in a damp
handkerchief at once. For unrolling rolls, or flattening out crushed
papyri, damping is needful. There is no need to steam them, as has
been done in museums. By dipping a towel or handkerchief in water, and
wringing it as dry as possible, there is enough moisture to penetrate
to a papyrus closely wrapped in it. If there were many turns then
carbolised water would be best, so as to avoid any decomposing during
a long penetration of the damp. Usually a single night is enough for
damping through half a dozen folds or turns, enough to render the
papyrus quite pliable. It can then be unrolled, or uncreased with the
fingers; and as each inch of it is laid flat it should be secured by
turning down newspaper or blotting-paper over it and sliding a board or
book over the flattened part. After leaving it between a dozen leaves
of paper to absorb the moisture for some days under pressure it is dry
and firm. Small pieces can well be carried in books, and larger sheets
in piles of paper between boards. When the papyrus is too rotted to be
damped, as the crossed layers of it would part, then it can only be cut
to pieces with a sharp penknife at every fold and turn; and each piece
fastened down on a sheet at once in place. This was the only possible
way to open the great Ptolemaic revenue papyrus over 40 feet long; even
a single turn of the roll needed to be cut into dozens of pieces.
For fastening down papyrus it is fatal to gum or paste it on to a sheet
of card, as the gradual contraction of the gum will break up the layers
of the papyrus. The safest way of all for very rotted papyri is to rub
a sheet of glass with beeswax, lay the papyrus on it, and press with
a warm hand until it sticks to the wax; then cover with another sheet
of glass. For ordinary firm papyri minute spots of paste, as small
as possible, should be put at every inch or two round the edges, and
farther apart in the middle; then a sheet of thin soft paper should
be pressed on it, to serve as a backing. Thus there is no wide space
pasted which can contract in future; and even if the papyrus has to
be remounted the paper can be torn to pieces behind it. The sheet of
mounting paper should be fixed under glass. But it is a mistake to
attach card to glass round the edges, as it bags away by damp and
warping, and leaves a large air space, which is very detrimental.
It is best to place the mounting paper between two sheets of glass;
or, for the sake of lightness and safety, the back may be of thin
picture-back-board, well baked dry, and free from cracks and knots. For
fastening the edges thin leather or linen may be glued around.
Dealing with carbonised papyri is an art in itself. So far as field
work goes the main work is to remove the earth entirely from the top
of the papyrus, so as to leave no weight upon it: then under-cut, and
take out the whole lump, with a block of earth under it. The papyri
must then, in the house, be carefully separated, one document from
another, by splitting apart and lifting with an ivory paper knife or
blunt table knife, the lighter the better, so as to feel the way with
it. Each separate roll should then be wrapped in soft paper (never
cotton-wool) and packed a few together, in small tin boxes. Thus they
will travel safely and without loss. The museum work is outside of our
scope; but broadly the Neapolitan plan of holding the pieces in place
with adhesive paper on the back is not so good as separate treatment of
each piece, laying it down in position on a sheet of glass with small
touches of paste, or perhaps pressing it on to waxed glass like the
rotted papyri. Burnt papyri are read by the difference of reflection of
the surface, and hence must be viewed with light from behind the eye,
or light reflected by a mirror placed almost between the eye and the
papyrus.
[Sidenote: Bead-work.]
Bead-work is often found in a state in which it cannot be moved owing
to rotting of the threads. Elaborate decoration with the winged scarab,
four genii, inscriptions, etc., is found on mummies of about the XXVth
Dynasty. But, if the threads are decayed, the beads are merely lying in
position, and will fall away if the mummy be tilted or shaken. In such
a case I have opened the wooden coffin very gently, cutting out the
pegs by which it was fastened. Having melted a pot of wax on a stove in
the tomb, I then dashed spoonfuls of it over the beads; it needs to be
thrown sharply, so as to splash out, or it runs off all in one line.
The wax must be only just barely liquid, or it will penetrate to below
the beads. When a sheet of wax is thus put over all the beads, the
sheet may be lifted up, and the pattern is seen in a clean condition,
reversed on the under side. The sheet can then be fixed with more wax
into a tray of wood, so as to keep it safely. If any of the beads are
not firm they can be heated and pressed farther into the wax. Strings
of beads are seldom found with the thread strong enough to hold
together. The earth should be loosened with a penknife, and blown away,
so as to disclose as long a line as possible, then the order of the
beads should be noted for restringing them, in the original pattern.
The tracing out and noting of a string of beads in a grave may often
occupy an hour or two hours, keeping the face close to the ground so as
to blew the dust away exactly, without disturbing the beads.
[Sidenote: Stucco.]
Stucco on wood we have already noticed, under the preservation of wood.
However firm the stucco may seem at first, the gradual contraction of
the wood will make it fall away; but when once saturated with paraffin
wax, this movement is stopped, and the stucco is held on to the basis.
Stucco on mud bricks is a difficult material to preserve. Three
instances may be given of dealing with it. Where the coat was a mere
whitewash on mud plastering, as at Tell el Amarna (Fig. 49), I removed
the bricks behind it by cutting them gently to pieces with a chisel;
thus the coat of mud plaster was left standing up a foot or more in
air, although it was entirely friable owing to white ants having eaten
out the straw from it. Then placing a box lid covered with sheets of
paper against the face, it was firmly grasped behind, and turned over
with the lid to support it, face down. Lying on the box lid it was
taken to the house; a frame of parallel bars of wood was made, each
an inch wide and an inch apart; each bar was coated with mud-and-sand
mortar, and then the frame was pressed gently on the back of the
fresco, and puddled in with mortar between the bars. On then reversing
the frame and box lid, the fresco was left resting on the frame,
with a bedding which was perfectly true, and incapable of warping or
contraction. To pack this a sheet of cotton wool was placed on the
face, a thin board cut to size placed over this, and string lashed
tightly round the face board and notches in the ends of the frame
bars. In this state it travelled quite safely, although the material
was so tender that a finger would push through it anywhere; this was
illustrated by a museum attendant at Cairo, when ordered to carry one
of the frames of fresco.
Where the stucco is thicker, about 1/16 inch, but wholly shattered
into minute chips, none over ¼ inch across, a different treatment was
necessary, as at Medum. The mass of plaster and stucco was laid face
down, the mud cut away behind it till about a square inch of shattered
plaster was bared at the back; this was covered with a thin coat of
fresh plaster (mixed in the palm of the hand); then another square
inch was bared and coated, and so on, until the whole of the mud was
removed and the old stucco all lay smeared with a thin coat of fresh
plaster on the back. A large slate was then cut to size; a pudding of
liquid plaster was poured on to the stucco and pressed out as thin as
could be with the slate. When it was set, the old painted stucco was
thus securely cemented on to the slate; light, tough, and portable, it
travelled to America in perfect state.
The third method is where the surfaces are curved. By cutting away the
back as thin as is safe, and setting in a firm backing of cement, even
this difficult subject may be dealt with, and removed safely.
[Sidenote: Gold.]
_Metals_ do not require much treatment in the field; but it is needful
to understand the condition of them in order to know how they can
be safely treated. Gold should be cleaned as little as possible, as
the old red surface is the best appearance of it; a little brushing
with camel-hair brush and plain water to remove the dust is generally
enough. Where there is much silver in it, as in electrum, the surface
is dark with chloride of silver; this may be removed with strong
ammonia or cyanide of potassium. Gold-foil often requires straightening
out into its former shape, but it must not be burnished in so doing, as
that expands the form.
[Sidenote: Silver.]
Silver is one of the most troublesome metals, as it is so very readily
attacked by chlorine and sulphur; and, moreover, it undergoes a
colloidal rearrangement by which it breaks readily into irregular
curved grains, and it is in this state as rotten as rotten brass. If
deeply corroded nothing can well be done to it; the lumpy crust shows
more of the original form than the metal would show if bared. When the
corrosion is but slight it may be removed, either by solution in strong
ammonia or cyanide of potassium, or by reduction. To bring the chloride
into the state of porous metal, it is only needful to place it with
zinc or iron in a solution of salt or weak vinegar or lemon juice, and
in a few hours the whole of the chlorine has gone over to the fresh
metal. The powdery silver left can be mainly brushed away in water, and
a little picking with a bone point will loosen it entirely. Of course,
the whole of the silver removed has come out of the body of the metal,
which is left porous and tender, although the face may be unbroken. It
will not bear, therefore, the same cleaning as new and strong metal. In
the case of silver coins in fine condition, each coin should be reduced
separately, and the whole of the old silver weighed with it before
cleaning it away, so as to recover the original weight. Silver must
never be put bare in a tin box, as the chlorine forms chloride of tin,
which deliquesces, and then attacks the iron and stains the silver with
brown rust. Often there is both chloride and lime on the surface, and
alternations of ammonia and weak acid are required for cleaning.
[Sidenote: Copper.]
Copper objects are distinguished from bronze by retaining usually their
pliability. This renders them much easier to clean, as they are seldom
deeply corroded, and the red oxide upon them will generally flake off
clean by blows, and leave the original face in perfect condition. A
very light hammer should be used, and sharp scaling blows be given, so
as to flake off even half-an-inch breadth of scale at once, without
ever touching the old face. In hollows which cannot so easily be
struck, an iron nail may be used as a punch, and struck so as to crush
the red oxide little by little. A copper object which scales freely is
a treat to clean, as the old face can be entirely bared, and appears of
a beautiful red-brown colour with all the detail quite perfect. Very
thin copper may, however, have entirely passed into green carbonate,
if buried in a damp soil; and in this case nothing can be done except
washing off the earth and dirt.
[Sidenote: Bronze.]
Bronze and brass need much more care than copper, as they contain a
mixture of alloys of very different oxidability; hence much of the
material all through the mass will have moved up to the surface and
been corroded there, while the form and size of the original may at
present contain only half the metal in a very porous and brittle
condition. In some cases bronzes may be scaled by blows like copper,
and they then appear in their best condition. But more often they are
too brittle, or the corrosion adheres too tightly, for it to be thus
removed. For cleaning off small quantities of green carbonate, vinegar
left to stand for some days does well. But the proper solvent of both
carbonate and oxide is dilute hydrochloric acid, about 1 to 10 or 20
of water, as this will not attack the metal, but only the corroded
parts. The objection to this solvent is that it leaves a thick mud of
white oxy-chloride of copper, which is difficult to brush off, and
which stains the skin green in handling. The treatment is to brush
off as much as can be easily removed, and then pickle in hyposulphite
of soda, which dissolves the white coat; if used hot and strong this
will clean the metal to a bright metallic condition. After all these
solutions, a long washing in many waters for two or three days is
needed to remove all trace of salts which might afterwards make further
corrosion. Minute traces of chlorides are specially dangerous, as
they decompose with carbonic acid in the air, forming carbonate, and
liberating the chlorine to attack more metal; thus a trace of chloride
will eat through any amount of copper. The extent to which bronzes
should be cleaned, should be ruled by the fullest display of original
workmanship: so long as more detail can be shown more crust should be
removed. But, if possible, some of the coat of red oxide should be left
on plain parts as a guarantee of the age of the work. To bare bronzes
entirely, and then oil and smoke them, is barbarous treatment, to be
seen in some museums. If something is desired over the bare metal, the
bronze may be left in a shallow pan of water, soaking for some weeks,
by which it will gain a tinge of red oxide over it which is suitable
and pleasing. Another mode of scaling is to heat the bronze over a fire
or in melted lead, and then plunge in cold water, which loosens the
scale from it. It often happens that a bronze has the original face
broken up by corrosion, and then no cleaning is of any use, the mass
of green carbonate shows more than any other surface would do. This
last and worst state is indicated by cracks in the outer coat, due to
further expansion of the inner body. A cracked bronze is best left
alone.
A frequent disease of bronzes is the formation of small granules of
translucent bright green rust. This is attributed to an organic growth,
which is infectious, and may spread through a collection. One of the
worst instances I dipped in carbolic acid, and this absolutely stopped
the attack, proving that it is not due to action of chlorine. But we
must not take this as a certain proof of the organic nature of the
mischief, in view of the inhibitory effect of anæsthetics, etc., in
stopping electric and chemical action.
[Sidenote: Lead.]
Lead is usually coated with white carbonate, the outer face of which
shows more than the metallic surface beneath. It should therefore be
let alone; but if it shows signs of further changes, due to salts in it
acting with damp, then soaking in several waters will probably make it
safe. If carbonate continued to be formed, I should try saturating with
paraffin wax.
[Sidenote: Iron.]
Iron can seldom be cleaned; but if it has only a little superficial
rust, this may be removed by placing it in the strongest nitric acid,
which dissolves the oxide but renders the iron passive. For ordinarily
rusted iron all that can be done is to arrest further changes. A
long soaking in water to remove all salts, and then baking dry and
saturating with wax, is a safe treatment and always available.
[Sidenote: Sorting.]
Sorting and joining fragments is sometimes very essential. In the
royal tombs of the Ist Dynasty we collected thousands of pieces of
stone bowls and vases. Only a very small number out of such cartloads
of fragments were of value as they lay; but so far as they could be
reconstructed they gave an important series of forms. To extract any
result it was needful to place together all the pieces that belonged to
each separate vase; and the same work frequently had to be done on a
lesser scale in dealing with groups of broken stone and pottery. Taking
the whole of the fragments which can be supposed by their position to
belong together, they are first sorted over for quality, making as many
divisions as are quite safe to be distinguished one from the other, so
that there shall be no chance of parts of one bowl being classed in
two different divisions. All the pieces of one division, sometimes as
many as 500 of one quality, are then to be laid out on tables,--the
pieces of brim placed at the top of the tables, and classed according
to form and curvature; the pieces of middle of the vase along the
middle of the table, all carefully laid with the axis vertical; the
pieces of base at the nearer edge of the table, classed according to
diameter. Taking then the first piece of brim, it is held at each end
of each other piece to which it can possibly belong; every possible
fit is thus found. Each piece of brim is to be thus tried with all
that follow it, those before it having been already tried with it.
When all the possible junctions of brim have been made, then a row
of joined brim pieces are to be laid on a board, and the angle which
each broken edge makes with the vertical is to be looked for among all
the broken sides of the middle pieces, looking for such slope at both
upper and lower sides if the tops are not distinguishable from the
bottoms of the pieces. Thus, say the first broken edge of brim slopes
at [Illustration] 20°, every piece broken at 20° [Illustration] or
[Illustration] must be compared to see if it will fit. At least twenty
different directions of fracture can be mentally distinguished, and
the slight curve and irregularities increase this to at least fifty
varieties, so that each piece of brim only needs actual touching with
about 2 per cent of the pieces of middle. When every possible fit
of brim to middle pieces is made, then the bases can be similarly
compared, having first fitted them by sorting the curvatures. A
load of 500 pieces will take several hours of this sorting, at the
end of which every possible fit will have been made. Not more than
half-an-hour or one hour at a time can be usefully given to such
sorting, as the eye and attention become too much fatigued to observe
the fits. When finished, all the fragments belonging to one bowl are
to be wrapped together, and a number given to the parcel; and the
odd pieces can be thrown away unless worth having singly. The method
for drawing the completed forms has been described in the chapter on
drawing.