|
Teaching >
Architectural Conservation II (HP 482) >
Dating of Old Houses
Henry C. Mercer, SC.D., Doylestown, Pennsylvania, 1923.
Wrought
Nails •
Cut Nails After 1800 • Hammer-Headed
Cut Nails • Stamp-Headed Nails •
Wrought-Iron Door Hinges • Cast-IRon
Door Hinges • Quirked, Ovolo Door Panels
• Machine-Made Door Panels • Door
Latches with Straight Lifts • The Norfolk
Latch • Blake's Cast-Iron Thumb Latch
• Pointless Wood Screws • Sawed
Laths • Conclusions
The following observations
are based upon notes taken upon the recent examination of about
one hundred and twenty old houses in Bucks county and Philadelphia,
Pennsylvania, built in the eighteenth and early nineteenth centuries,
and it seems probable that the conclusions apply not only to old
dwellings in Pennsylvania, but also to those in New York, New
England, and the Southern states, where the same builders' material,
carpenters' methods, tools and hardware were used during the period
in question. The conclusions are as follows: that old houses may
be dated within reasonable limits by the nails used; the hinges;
the door panels; the wrought-iron thumb-latches; the Norfolk latches;
the cast-iron thumb-latches; the wood-screws; and the sawed laths.
Wrought
Nails
Handmade
(wrought) nails of soft malleable iron, with rectangular shanks,
drawn by hammer blows to a point and with clearly hammer-marked
heads, were from time immemorial, universally used in house
building, until about 1800 (in Philadelphia, 1797) when cut
nails, because of their much greater cheapness, everywhere immediately
superseded them. Therefore, where the original nails of a house
are wrought, the house dates before about 1800; or, where cut,
vice versa, after that date. All the evidence examined establishes
this fact, with the following exceptions; namely, that long
after 1800, wrought nails, to stand the jar, and because they
would clench, continued to be used in the facings of window
shutters; in the battens of doors; in the overlap of boards
(old style) in lathed room partitions: or on door latches, etc.,
until about 1850. But these exceptions are not typical of the
nails used to build houses after 1800. Nails used at the time
a house was built are nearly always to be found in the garret
floors. The wrought nail, no matter what its size, as generally
used in house construction, is easily distinguished from the
machine-made nails, called cut nails, above re-furred to, and
described later. It was made from rectangular strips of malleable
iron, several feet long, and about a quarter of an inch thick,
called nail rods, which were furnished to the black-smith or
nailer, who, holding one of them in one hand, heated its end
in his forge, and then, on the anvil, pointed it with the hammer
on all four sides. Next, he partly cut it, above the point,
on the "hardy," with a hammer blow, and then, inserting
the hot point into the swage hole, of his so-called ²'heading
tool,' he broke off the rod and hammered the projecting end
so as to spread it around the top of the hole; after which,
the cooling, shrunken nail was easily knocked out of the orifice.
Wrought nails, as free-hand forged products, vary greatly in
style and shape, but the evidence examined has not as yet furnished
any definite elate for any of their variations.
Cut
Nails After 1800
The far more easily
made cut nail, as the evidence clearly shows, consists of a
rectangular, tapering shank of iron, not hammered into a point
by hand, but tapered, by a single cut, across a plate of iron.
The smith was here furnished, not with a nail rod, but with
a strip of plate iron, several feet long, about two and a quarter
inches wide, and often about one-eighth of an inch thick. This
strip he slid into a cutter, worked at first by hand power,
resembling those used by bookbinders to trim books, and not
here shown. This cutter, rising and falling rapidly, clipped
off the end of the iron plate crosswise into narrow, tapering,
rectangular slices or nails, whose length was established by
the width, and thickness, by the depth of the nail plate. The
taper of the cut alone, produced the point, but not the head.
This was made at first by dropping the freshly cut piece, point
downward, into a slotted clamp or vise, and then spreading the
larger projecting end with a hammer, as in the case of the wrought
nail. Cut nails are easily distinguishable from wrought nails
by the following very apparent differences. Both have rectangular
shanks, but the wrought nail tapers on all four sides; the cut
nail, on only two opposing sides; the latter nail being as thick
(namely the thickness of the nail plate from which it was cut)
at the point as at the head. Moreover, the two cut sides of
the cut nail show very plainly, minute parallel striations,
always absent on the wrought nail, marking the down smear of
the cutter. The evidence conclusively shows that these cut nails
every-where superseded the ancient wrought nail at the end of
the eighteenth century, namely, not long after 1797, when two
cut-nail factories had been established in Philadelphia, and
there-fore, if used by the builder, they will date a house as
having been built after that year.
Hammer-Headed
Nails c. 1800 to c. 1825.
A still
further examination of cut nails, from dated houses, shows that
they may be distinguished into two classes; namely those appearing
between c. 1800 and c. 1825, with imperfect or irregular heads,
or, more particularly, hammered heads; that is, heads showing
the facets of more than one hammer blow, and those appearing
after c. 1825, and throughout the following century, with stamped
heads, showing level tops impressed by a single blow or stamp.
Information gathered with difficulty from the Patent Office
records and books, makes it probable (subject to correction
by dated nails) that in general, up to 1825, the nail-cutting
machines had not been perfected; in other words, that while
after 1825, nail machinery produced cut nails at a single operation,
before that time, two machines, run by hand power, but not yet
by steam, nor even by water, one to cut, as described above,
and another, probably nothing more than a special vise to hold
the shank while hand-hammering the head, were used in the manufacture
of cut nails. The hand-cranked machine, for cutting and heading
nails at one operation, patented by Nathan Read of Salem, Mass.,
in 1798 (See model at Essex Institute, Salem, Mass.), was not
a success. Neither were any of the other ²cutting and heading"
machines, or simple ²heading" machines, in existence or patented
at that time, as is shown by the evidence of the nails themselves,
and further in the Diary of Rev. William Bentley, who visited
Read's nail works in 1810 (See Essex Institute Historical Col-lections,
April, 1918, page 113), and found that the workmen were then
heading nails in the only way thus far successful, namely, by
hand, "as it is found heading is done better by hand than by
any machine as yet invented both as to time and good-ness of
execution. Joseph Whitaker (See his manuscript diary in the
library of the Bucks County Historical Society) was also thus
making cut nails in Philadelphia, from 1809 to 1816-20, by a
double operation; namely, cutting the plates with a hand-cranked
ma-chine and afterwards hammer-heading the shanks held in a
clamp worked by a foot lever. It further appears, that, at first,
since the knife of the cutting machine was set diagonally so
as to cross-cut the nail-plate into a tapered slice, the workmen
had to turn the plate upside down at each stroke, so as to continue
the taper by reversing the cut; and the very earliest cut nails
(1800 to c. 1810) prove this fact by the down smear of the knife,
round-edged above and sharp below, being reversed on the two
opposing cut sides of the nail shank. They also show, that very
early in the nineteenth century, this troublesome turning of
the nail-plate was superseded by wriggling or staggering the
blade of the cutter during the operation, so as to reverse the
taper at each stroke without, turning the nail-plate. At first,
also, in order to dispense with the difficulty of the usual
heading, angle-headed (L headed) and headless nails called "brads"
were made. But as these latter continued in use for certain
purposes (often for floors) until long after the middle of the
nineteenth century, their confused evidence should here be thrown
out of consideration.
Stamp-Headed
Nails After c.1825
An examination,
not only of the records above mentioned but also of dated nails,
shows that about the year 1825, the cut-nail machine, still
working by water-power rather than by hand, and not yet by steam,
had been so perfected as to make cut nails no longer by two
operations but by a single operation in one ma-chine, in which
the apparatus cut the nail, instantly clamped it and, at a single
blow, stamped the head. These stamped heads, at first (c. 1825
to 1830) comparatively thin, lopsided and imperfect, became
more thick, square and typically regular after 1830 and are
always easily recognizable after about 1840. But regardless
of their variations, in any case, stamp-headed cut nails, if
used in constructing a house, reasonably date it as after about
1825.
Wrought-Iron
Door Hinges
The
evidence clearly shows that in the Colonial period in America
the common iron, house-door hinges were made always of wrought
iron until 1776 to 1783, when cast-iron hinges suddenly and
universally took their place. The old wrought hinges appear
in two common varieties in the houses examined; namely, the
so-called H or HL hinge, cut out of heavy sheet iron and fastened
against the face of the door with screws or clenched wrought
nails, or the ²strap" or ²hook and eyeÓ hinge; namely, a long
strap, bolted, riveted or nailed with clenched nails, against
the door and turning on a hook or gudgeon which latter was either
spiked into the lintel, or, where the lintel was too thin for
spiking, set upon a plate, variously shaped, and sometimes strengthened
with a projection or prop called a ²rattail." While the H and
HL hinges (many of which were probably factory-made and imported
from England) and nearly all of the strap-hinges, were found
plain, a few of the latter, by no means typical and generally
over-exhibited in museums, show floriated decorations. It further
appears that hand-made, wrought-strap hinges (still common in
1923 on barn doors in eastern Pennsylvania and elsewhere), continued
to be used on outer house doors and window shutters, long after
1783, and hence, when so found, should be disregarded as proof
of dates, But with these exceptions, the evidence abundantly
shows, that where wrought hinges (generally HL, more rarely
strap) are found on original inner house doors, they date the
house as Colonial, or built before the Revolution.
Cast-Iron
Door Hinges
Cast-iron
door hinges, called butt hinges, comparatively small, compact,
book-shaped, mortised into the edges, not set upon the faces
of the door, of the common present type (See Fig. 8), be-cause
of their superior cheapness, came into universal use, no less
suddenly, though a little earlier, than cut nails. They were
invented in England by Izon & Whitehurst, and patented by British
patent No. 1102, October 3, 1775, and were at first imported.
After the interruption of British trade and house building by
the Revolutionary War, they everywhere superseded the old wrought
hinges, about 1784, after which they appear with- out significant
exception, on all the dated houses examined by the writer. Hinges
of this shape and name, i. e. butt hinges, of wrought-iron or
brass, and never of cast-iron, bad been made before 1775, generally
for closets, or furniture, but none was found by the writer
on room doors, in the houses examined. Cast-iron butt hinges
also show differences and improvements in construction (not
studied closely) after about 1800. But regardless of these variations
and allowing for the above noted survival of wrought strap hinges
on outer doors and shutters, these cast butt hinges, found upon
the original doors of houses, will date the latter as post Colonial
or built after c. 1776-1783. This examination of old houses
has shown no more remark-able and unlooked-for fact than that
the door panels, before c. 1776, if edged as usual with mouldings,
always show a plain. i.e. unbeaded ovolo or quarter-round molding
on their outer margin, while immediately following the Revolution,
after 1783, these same ovolo mouldings become scored with one
or two quirks or beadings, or change into the ogee. It seems
probable that this observation will apply not only to door and
shutter panels, but also to wall and furniture panels. Nevertheless,
lacking sufficient information, as yet, we here limit it to
doors where it is significant enough. More probably caused by
some technical change or improvement in journey, not yet explained,
than by mere fashion, this sudden, marked and universal change
in door panels seems all the more surprising, since beaded or
quirked ovolo and ogee mouldings appear elsewhere in the woodwork
of old houses, as, for instance, in cornices and the framework
of mantels. Further, since old carpentersØ books describe hand-planes
used to produce the latter mouldings considerably before 1776,
it would seem reasonable to expect to find some exceptions to
this rule; but the evidence of the houses in question shows
none in the region examined, so that, subject to future correction,
the information thus far gathered shows that hand-made door
panels with plain ovolo frame work, if part of the original
construction, will at once date a house as Colonial, or as built
before c. 1776.
Quirked-Ovolo
Door Panels, from c. 1776 to c. 1835.
As above
stated, the evidence gathered shows that after c. 1776, door
or shutter panels, in which the outer frame consists of an ovolo
moulding with one or two beads or quirks (See Fig. 10), or an
ogee, suddenly and universally supersede the old plain ovolo
moulding, described as previously used, and continue in use
on doors and shutters until machine-made mouldings take their
place about 1835 (See Fig. 11). In all the old houses examined,
no significant exceptions to this rule, or extended survivals
of old, plain ovolo panels, during the period in question, have
been found, so that thus far, the evidence abundantly shows
that the more ornate (i.e. beaded or quirked ovolo) door panels
described, if part of the original construction of a house,
will date it as built between c. 1776 and c, 1835.
Machine-Made
Door Panels After c.1835
Besides
the two significant changes in door panels, above noted, a third
change, later but no less marked, took place in their construction
upon the general introduction of wood-working machinery, wood-planing
mills, etc., about 1835. Revolutionary machines, of immense
importance, to plane boards, make mouldings and otherwise work
wood, had been in-vented in England by General Bentham, just
before 1800 (See Knight's American Mechanical Dictionary) and
no doubt were introduced into the United States and used about
Boston, New York, Philadelphia, etc., between 1790 and 1835.
Hence, very early machine-made door panels may be found later,
in these and other old American cities, to prove the fact. But,
in any case, these woodworking machines would have been run
very restrictedly by water-power and not by steam, and the evidence
shows that they were not established or their products used
in the Pennsylvania country until after the general introduction
of steam-power which gave birth to the modern factory about
1835. Before that time, in the houses examined, all mouldings
on door panels, whether of the plain or quirked ovolo or ogee
type, above described, were hand-made and appear as solid parts
of the panel, planed by hand-moulding planes upon its framework;
while after that time they were machine-made and nailed on,
as loose strips, around the sunken outer marginal recess of
each panel. It is not necessary for this purpose to consider
the various sizes and shapes of these machine-made mouldings,
nor to reason from the fact that they were introduced, not suddenly,
but gradually, that the old styles of hand-made panels continued
in use for a good while after their introduction. To discover
that loose strips of moulding have been nailed on around the
sunken outer marginal recess of a panel is sufficient: that
fact, where they are part of the original house construction,
establishes the date of the house as not earlier than about
1835.
Door
Latches with Straigh Lifts, Before 1800
Besides
other door fastenings, namely box knob locks, wooden latches,
brass latches, German lever latches, boxed or unboxed, knob
latches, etc. not here described, many original doors in old
houses still standing, show their original wrought-iron thumb-latches,
made of malleable iron by blacksmiths in five hammered pieces,
i.e. the hand grasp, an lever with thumb press at one end penetrating
the door to raise the bar; the bar thus lifted; the staple holding
the bar against the door face; and the catch, a "figure 4" shaped,
notched, iron piece, spiked into the lintel of the door, into
which the bar falls. These old latches are sometimes decorated,
but commonly plain, sometimes home-made and sometimes probably
imported. Some-times they show their thumb-lifts fixed on swivels
(the swivel-lift latch); sometimes the thumb-lifts are notched
into holes (the perforated cusp latch), and sometimes their
latch-bars appear with, but generally without, a knob or curl
or pinch grasp. As yet no fixed types have been found to which
dates may be ascribed beyond the following; namely, that the
inner end of the lift, opposite the thumb-piece, commonly though
not always appears straight before about 1800; after which it
more and more often shows the familiar down curve under the
bar, characteristic of modern cast-iron latches. Doors latched
with these straight-lift latches, some of which are very short,
are sometimes hard to open, and sometimes, as if to remedy the
difficulty, knobs or pulls appear on the bars of latches of
early Colonial date. But these early knobbed-bars are rare and
it seems all the more remarkable that the very helpful down-curve
above mentioned should not have been more generally used be-for
1800; nevertheless curved latch-lifts have been heard of by
me, and seen by Mr. Frank K. Swain, in old houses in England,
and in Pennsylvania, dating from the earlier period in question,
e.g., several at the Community House, Bethlehem, Pa., built
about 1742, and several at the Letitia Penn House, Philadelphia,
c. 1682 (doubtful). Since the writing and first publication
of this paper in Old Time New England, The Bulletin of the Society
for the Preservation of New England Antiquities, for April,
1924, Mr. Albert H, Sonn has seen a curved lift-latch on a library
door at Hadham, Conn., traced to an old mill built about 1740;
one on a house at West Stockbridge, Mass., and one at Newfane,
Vt., besides finding more recently a dozen or more in various
parts of the eastern United States. Dr. A. Bertram Gilliland
has also found several with scrolled, upturned lifts in the
Stebbins House at Deerfield. Mass., built in 1772; one from
the Pastor Williams House, Deerfield, built in 1770, and one
at Washington's Head-quarters, Newburg, N. Y., built before
1800. If more should appear later, the present evidence shows
that they will continue to occur 'as exceptions, and that in
general a down turned latch-lift, if part of the original construction,
will date a house after 1800.
The
Norfolk Latch, After 1800
The
very conspicuous Norfolk latch, is easily distinguished from
the wrought thumb-latches, in having its hand-grasp not enlarged
at each end into plates, or cusps, but riveted upon a long,
narrow, sheetiron escutcheon. Though long known in England as
hand-wrought by local blacksmiths, it nevertheless appears in
the American houses examined, as a factory-made and not smith-wrought
product probably at first imported from England. Gradually taking
the place about 1820 of the other forms of thumb-latch and competing
with the knob-latch and the German lever latch, it rivals, for
a while, the newly invented earthen door-knob with cast-iron
box, until it is generally superseded by the latter and by Blake's
patent cast-iron thumb-latch of 1840. The evidence shows that
these factory-made Norfolk latches were constructed sometimes
with, and sometimes without, a knob on the bar; sometimes, at
first, with a straight lift and sometimes, later, with a curved
lift, sometimes, at first with a spiked catch and sometimes,
later with a catch perforating or riveted upon a plate. But
without attempting to infer too much from these variations,
we may at least conclude, from the evidence, that the factory-made
Norfolk latch, if contemporaneous with the building, will date
a house between 1800 and 1840, or, allowing for survivals, 1850.
Blake's
Cast-Iron Thumb Latch, After1840
Numerous
dated examples found, show that Blake's typical cast-iron thumb-latch,
with circular catch-plate mortised and screwed into the door
lintel, hollow patent bar-pivot, hollow staple guard, and saucer
lift with opposite down-curve, patented by United States patent
No. 1704, July 21, 1840, first came into general use on and
after that year. It seems probable that this latch was preceded
by rare cast-iron experiments or improvements, i.e. cast-iron
grasps on older wrought latches of the Fig. 12 type, etc., and
was closely followed by evasive copies or patent infringements.
But Blake's latch was, and still is, (1923) the cast-iron latch
par excellence, and without concerning ourselves with earlier
un-patented predecessors or variations of it or copies or patent
infringements of its very typical catch or staple, this latch,
when complete and original, as the evidence clearly shows, will
date a house built after 1840.
Pointless
Wood Screws Before 1846
The
unmistakable pointed wood screw, now universally used, was patented
by United States patent No. 4704, August 20, 1846, before which
time, all wood screws in general use, unless pointed by hand-filing,
were blunt. Because these pointless screws would not start by
driving into the wood, penetrate, except by a previous gimlet
or brad-aw hole, the pointed wood-screw suddenly and universally
superseded them. Therefore, the wood-screw if pointless and
original, will date a house before 1846; if pointed, after that
date. These facts, marking the end of the old house-building
period, though only applicable to the very latest buildings,
are nevertheless important, since they may help to detect wholesale
restorations or additions and show when kitchen fire-place doors
stopped open-fire cooking, or where old latches, hinges, or
doors have been shifted out of time or place.
Sawed
Laths, After c. 1825 to 1835
Sawed
laths, i.e. thin strips of machine-sawed wood, about three feet
long, by two inches wide, by a quarter of an inch thick, as
keys for interior wall furring and partition plastering, first
appear about 1825 to 1835. Though sawed, they were not produced
by the water-run, vertical-frame saw of the old saw mills, but
were first made by circular saws, abou t1825 to 1835, on the
general introduction of the circular saw, before which time,
riven laths, i.e. hand-split with a frow and mallet, were invariably
used, and no such thing as a sawed lath existod. Riven laths
were occasionally made and used for some time after the introduction
of sawed laths, and therefore will not date a house as built
before 1825, while sawed laths will, if original, date it as
built after that time.
Conclusions
In conclusion,
it should be said in general, that in collecting and attempting
tc estimate the above facts, it soon became certain that very
few of the old houses examined had escaped alterations and repairs
and therefore, unless the details above noted could be proved
in each case to be part of the original construction, their
evidence only led to error and confusion. With this reservation,
reasonable certainty was always sought for and often found.
Very few houses appeared to have been raised or broadened. Therefore
their original garret floors remained intact and the conclusive
evidence of nails used therein, was easiest reached. When rarely,
because of new floors, or L headed cut nails. this failed, we
generally found it on staircases, in wash-boards or else-where
in the house, and when, at times, this evidence seemed contradictory,
some Further fact, family tradition or historical record, showed
that old doors or hinges, screws, or latches, had been inserted
out of date into new houses, or vice versa. Doors appeared original
if set in original partitions; if frequently duplicated; if
not cut down on their margins: and if with their hinges not
covering old mortise nail or screw holes or outlines of removed
hinges. Door panels; if on original doors; if frequently repeated
or matching shutter panels. Latches; if often duplicated, and
not betrayed as resettings by the marks of nail, screw or lift
holes, etc., or of other door fastenings. Pointed or pointless
wood screws; by their general use or appearance with otherwise
original wood or iron work; and sawed laths; by their original
use in partitions or in original furrings over rough unplastered
walls. Out of at least one hundred and fifty houses examined,
about fifty were found dated by documentary evidence, or by
date-plates or wall-stones; and (he evidence of nails, woodwork
and hardware, first studied in these dated buildings, always
repeated and never contradicted itself in the undated houses
examined later. As far as this evidence goes, it is very positive;
but as yet, though quite definite after the Revolution, it fails
to fix any subdivisions of time for the Colonial period (1650
to 1776).
|
