88
or knotting a very smooth surface may
be obtained. The Chinese do some ad-
mirable works in this style, and the figures
we see in slight relief on their works,
and which are occasionally seen in some of
the oldmonkish missals, may be done by using
a mixture of parchment size, fine whiting,
pipeclay, and a little honey. If these are well
mixed together to aconsistency so that it may
be used with a pencil, and the figure or orna-
ment to be in relief is coated over three or
four times with this mixture, the one coat
being dry and hard before the next is put on,
we shall produce the same effect as the
Chinese do. The degree of relief will of
course depend upon the number of coats of
the mixture which is put on. The work
must now be smoothed with the sandpaper,
well dusted, and sized with the parchment
size, then with the oil-size and bronzed or gilt
in the usual manner. If care and skill are
used in putting on the filling-up mixture,
the inequality of the surface may be made to
correspond with the projections of the figure
or drapery, which will assist the effect of
the work by adding a slight portion of light
and shade to the relief.
(To be continued.)
>
HAMMER-BEAM ROOFS.
(Concluded from page 43.)
N the case, however, of roofs of higher pitch
where the arched brace had been straighter,
the collar being carried higher (Fig. 12), it had
stood much better. Many of the old hammer-beam
roofs showed a greater regard to the direct action
of the timbers by way of a brace. This was
eonspicuous, amongst numerous similar examples, in
the roof of a building at Islington, the sketch of
which (Fig. 13) is taken from ‘ Smith’s Ancient
Carpentry,” and also Fig. 14, from Netley, in Hamp-
shire, which showed a row of 14 pins on the collar
tenons, and 8 on the rafters. The roof of the hall
at Lambeth Palace, Fig. 15, appeared to follow ex-
aetly the structural principles of Westminster Hall,
but the arch is semicircular instead of pointed, and
the queen and collar cross each other, so as to give
an additional bearing to the rafter, and an additional
breaking strain. In the roof at Eltham Palace the
arched rib is continued down more nearly to the foot
of the principal, thus increasing to a great extent the
tensile strain by bringing its line more directly into
the line of the principal, and neutralising the com-
pressive strain on the back of the archedrib. Take
a common framing (Fig. 16); a b c are in compres-
sion, and d is in tension. If unequally loaded, as at
e, a strut e f will contribute considerable support.
But this will be by so much, merely taking load from
aand b and throwing it on f. The load may be
again distributed, however, by adding a tension f, g,
and it will be seen that the reverse of this, viz., a
thrust at gf ora tension at fe, would be useless with
the load at e. Now, if another equal load be applied
at g the same process will apply in the reversed order,
and hence to allow of the possible contingency of un-
equal load, whether by pressure of wind or otherwise,
we require such struts as should act as tension-pieces
when under the extra strain from either direction.
And the strength required for tie at bottom and for
these fillings is very little as compared with the com-
pressive strength for a,b, andc. Extending the
operation of this truss by carrying the struts beyond
the tie or by bending the tie to embrace them, thus
converting the tie into a rigid arch, it will be seen how
much this adds to its strength and rigidity.
(Fig. 17.) Supposing the curved tie to be of pro-
portionate strength, and the whole framed together
if need be by numerous subsidiary trusses and
braces. (Fig. 18.) It is plain we have a truss
true in principle and construction. The curved
tie is not likely to sag and straighten, for it is
slung by its king-post, which again is carried by
the struts framed into the extremity of the tie-
beam, and the straightening of this curved tie is
manifestly the only contingency by which the truss
could sag, supposing the joints and the framing to
be of corresponding strength. We will now vary a
little the proportions and arrangements of the various
parts, and although the pressure is not distributed
so equally between the two sides, yet its structural
principle is scarcely changed; one end is more
rigidly trussed than the other—the end, namely,
which has to bear the greater strain. Tor if we set
up two such trusses against each other, and unite
them by bringing the two last pieces together so as
to become, as it were, one king-post, and the two j
THE BUILDING NEWS. Fes. 2, 1872. next into a line forming one collar, we get not only two rigid principal rafter trusses, but we get them so united that two of their main ties are common to each other, and thus afford immense mutual support. Then, in addition to this, the two great braces come together, and besides their office of bracing severally the two trusses, they become a mighty arch, acting as a counter-support to the actual weight of the timber itself. It is scarcely needful to add that such is the construction of the roof of Westminster Hall. It will be seen that there is a cross-piece from wall to rafter, and also a diagonal timber running through the thickness of the tracery in the spandrel at the back of the arch. This appears to be a strut, as it has an abutment prepared for it at top and bottom, and it forms one side of a truss common to the arch and hammer- beam on the one side, and to principal rafter and hammer-beam on the other, so as to prevent the racking of the otherwise rhomboidal spandrel. It is very possible that the large arch, by its mere gravity, helps to counteract the spreading of the foot of the principal as well as to prevent the spreading of the angle between the queen-post and the main- collar. Mr. White said he was sorry to have to differ so radically from Mr. Morris, who had brought much intelligence and earnestness to bear upon the subject of constructive carpentry, but he could not let some of his conclusions go unquestioned. Mr. Morris observed, on page 15 of his work, upon the imperfect degree in which certain properties of timber and the principles of constructive carpentry were understood, though the principles of masonic construc- tion were mastered by our ancestral architects. In page 18 he shows that arches or curved pieces in timber-work were always used solely by way of support—repulsion, compression, or passive resist- ance—in Medieval work. Allowing for the alleged ignorance or want of theoretical construction, it seemed tolerably plain that these timbers, whether the architects knew it or not, did very often act by way of tying or bracing to an extent which it would have been dangerous tohave dispensed with. If, however, Mr. Morris meant by passive resistance —which Mr. White did not gather from his mode of introducing it—that the forces were perfectly at rest, whether through means of counterpoise or counter— erection of the tensile and compressing forces, Mr- White would not find fault, for it was the exact adjustment of these forces that gave rigidity to a truss of any description which was likely to be sub— jected to unequal pressure. It was, however, per— fectly true that the general system followed in roofs of all sorts was rather superposition than suspension by way of trussing. Especially was this so in the case of long ranges of framed or coupled rafters, whether with or without the massive tie-beams. The framing of these rafters, however, showed a system of truss— ing in one sense. It was in the highest degree one of bracing equally witb strutting, and these large horizontal untrussed beams laid across the building in order to the erection of posts to support the collars, so common in early Medizval roofs, acted without doubt, not only for the purpose of such support, but also as horizontal ties to the wall— plates and walls to such an extent as actually to save from ruin many a building which otherwise would have suffered irremediably, whilst others have gone to destruction from their decay or removal. The super-refined trussing introduced in the last century, and developed to an alarming extent at the present day in order to reduce to a minimum the amount of timber to be used in a work—a work which was to present an appearance as well as reality of strength—was, doubtless, never for a moment present to their minds. It would have been unworthy of their high aims. We should have had now, indeed, no chance of entering into such a dis- cussion as this. The very slightness of the scantling- and bulk of timbers now commonly used, as com- pared with what they used, would have left us but scant fragments of the past. But Mr. White said he should be sorry to argue from this that they necessarily had no knowledge of, and never employed timbers by way of bracing and tying, and that their theory, as well as their practice, was stricily con- fined to that of supporting and strutting. The type of construction found in the roof of Trunch Church, Norfolk, as figured in Brandon’s “Timber Roofs” might seem to countenance the superstructural view (see Fig. 19). But, indeed, it scarcely applies, both on account of the smallness of the span (19 feet) and the absence of this arch. And so with many others. Moreover, this roof does show bracing by means of the key-shaped king at the top of the FIGAS
ZAP arched principal. In the type shown by the roofs of
Cj the Chapel of S. Mary’s, Suffolk (Mig, 20.), and
G3 ZY Pulham, Norfolk (Fig, 21), it is evident the great
Gj GG aim was to create a kind of bite or grip by leverage:
Qa BA of a very opposite character, the weight being,
