A R C H I T E 
-plan ; it is required to find the angle rib, and the form of 
the boards that will cover it when the ribs are fitted up. 
Let B in the figure be the given rib ; divide the curve into 
any number of equal parts, as four, and lay them-at D 
from a to 4, which bifefls b b, the fide of'the polygon, at 
right angles ; through thefe points draw lines parallel to 
the fide b b of the polygon ; at B and D make iratD equal 
to c c, between B and C 2 d equal to ,dd, and 3 e equal to 
e e, See. and through the points b, c, d, e,f draw a curve 
line, which will be the form of the boarding ; from the 
points g,f, e, d, c, draw lines at right angles to g b, the 
bafe of the angle rib, and trace the rib C from D as they 
are marked by the letters, which is plain. The more parts 
there are in this operation„the truer it will be, or any other 
of this nature; and in the fame manner may the covering 
and angle ribs of any other polygon be found, whatever 
may be the form of the ribs. 
With refpetft to the timbers employed in roofing, many 
opinions have been offered by different architects. It feems 
generally agreed, that all beams of tw-enty feet and up¬ 
wards ought to be truffed up in one or more places. 
Beams fliould never exceed fifteen feet in bearing, nor raf¬ 
ters more than ten, efpecially in roofs of very low pitch, 
whofe coverings have a much greater prelfure on their 
rafters than thofe of higher pitches. It fhould alfo be 
remembered, that timber is much weakened by its own 
weight, unlefs it (land perpendicular; and that, whenever 
a mortice is to be cut in the fide of a piece of timber, it 
will be much lefs weakened when cut near the top, than if 
cut near the bottom, provided the tenon be driven in hard 
to fill up the mortice. The bending of timber will al¬ 
ways be nearly in proportion to the weight that is laid up¬ 
on it; no beam ought to be trailed for any length of time 
with above one-third or one-fourth of the weight it will 
abfolutely carry ; for experiment proves, that a far lefs 
weight will break a piece of timber when hung to it for 
any confiderable time, than what is fufficient to break it 
when firft applied. 
If feveral pieces of timber of the fame fcantling and 
length are applied one above another, and fupported by 
props at each end, they will be no fironger than if they 
were laid fide by fide; or, which is the fame thing, the 
pieces that are applied one above another are no (Ironger 
than one fingle piece, whofe width is the width of the fe¬ 
veral pieces collected into one, and its depth the depth of 
one of the pieces; it is therefore ufelefs to cut a piece of 
timber lengthways, and apply the pieces fo cut one above 
•another, for thefe pieces are not fo ftrong as before, even 
if bolted. For, fuppofe a girder 16 inches deep, and 12 
inches thick, the length is immaterial, and the depth be 
cut lengthways in two equal pieces ; then will each piece 
be 8 inches deep, and 12 inches thick. Now, according to 
the rule of proportioning timber, the fquare of 16 inches, 
that is, the depth before it was cut, is 256, and the fquare 
of 8 inches is 64; but twice 64 is only 128, therefore it 
appears that the two pieces applied one above another are 
but half the iirength of the folid piece, becaufe 256 is 
double 128. 
Thefe obfervations will be found of ufe, when timber is 
wanted to fupport a great weight; for, by knowing the 
fuperincumbent weight, the firength may be computed to 
a great degree of exadlnefs, lo that it (hall be able to fup¬ 
port the weight required. The confequence is almolt as 
bad when there is too much timber, as when there is too 
little ; for nothing is more requifite than a juft proportion 
throughout the whole, in order that the ftrength of every 
part may always be in proportion to the ftrefs ; for, when 
there is more ftrength given to fome pieces than to others, 
it encumbers the building, and confequently the founda¬ 
tions are lefs capable of fupporting the fuperftrudhire. 
No judicious perlon, who has the care of conducting build¬ 
ings, fhould therefore rely on tables of fcantlings, fitch 
as are commonly found in books; for example, in ftory 
polls the fcantlings, according to feveral authors, are as 
follow : 
C T U R E. u s 
For 9 feet high 6 inches fquare 
12 - 8 
35-10 
iS - 12, See . 
Now, accoi-ding to this table, the fcantlings are increafed 
in proportion to the height; but there is no propriety in 
this, tor each of thefe w ill bear weight in proportion to 
the numbers 9, 16, 25, and 36, that is, in-proportion to the 
fquare of their heights, 36 being 4 times 9; therefore the 
piece that is 18 feet long will bear four times as much 
weight as that piece which is 9 feet long; but the 9 feet 
piece may have a much greater weight to carry than an 
18 feet piece, fuppofe double, in this cafe it muft be near 
12 inches fquare inftead of 6. The fame is alfo to be ob- 
ferved in breaft-fummers, and in floors where they are 
deftined to fupport a great weight; but in common build¬ 
ings, where there are only cuftomary weights to fupport, the 
common tables for floors will be near enough for practice. 
In choofing timber, care fhould be taken to avoid knotty 
and crofs-grained wood, See. fuch pieces being not lo 
ftrong as thofe which are ftraight in the grain ; and, if care 
be not taken in /electing the timber for a building, all 
rules that can be laid down will be ufelefs and nugatory 
in regard to its ftrength. Let it alfo be remembered, that 
in all timber there is moifture, wherefore all bearing tim¬ 
ber ought to have a moderate camber or roundnefs on the 
upper fide, for till that moifture is dried out the timber 
will fwag with its own weight. And obferve, that it is 
beft to trufs girders when they are frefti fawn out, for by 
their drying and Ihrinking, the trulles become more and 
more tight. Alfo let all beams or ties be cut, or in fra¬ 
ming forced to a rounding fuch as an inch in twenty feet 
length, and that principal rafters alfo be cut or forced in 
framing^as before ; becaufe all joifts, though ever fo well 
framed, by the fhrinking of the timber and weight of the 
covering will fwag, fometimes fo much as not only to be 
vifible, but to offend the eye ; by this precaution the trufs 
will always appear well. Likewife obferve, that all cafe 
bays, either in floors or roofs, do not exceed twelve feet 
if poflible; that is, do not let the joifts in floors exceed 
twelve feet, nor the purlines in roofs, &c. but rather let 
their bearing be" eight, nine, or ten, feet ; this fliould be 
regarded in forming the plan. In bridging floors, do not 
place the binding or ftrong joift above three, four, or five, 
feet apart, nor the bridging or common joifts above ten or 
twelve inches apart, that is, between one joift and another. 
Alfo, in fitting down tie beams upon the wall plates, never 
make the cocking too large, nor too near the outfide of the 
wall plate ; for the grain of the wood being cut acrofs in 
the tie-beam, the piece that remains upon its end will be 
apt to fplit oft', but keeping it near the infide will tend to 
fecure it. And obferve further, never to make double 
tenons for bearing ufes, fuch as binding joifts, common 
joifts, or purlines ; for, in the firft place, it very much 
weakens whatever it is framed into, and in the fecond place, 
it is very rare to have a draught in both tenons, that is, 
to draw both joints clofe ; for the pin in palling through 
both tenons, if there is a draught in each, will bend fo 
much that, unlefs it is as tough as wire, it will break in 
driving, and confequently do more hurt titan good. Roofs 
will be much ftronger if the purlines are notched above 
the principal rafters, than if they are framed into the fide 
of the principals ; for by this means, when any weight is 
applied in the middle of the purline, it cannot bend, being 
confined by the other rafters; and, if it do, the lidesof the 
other rafters muft neceffarily bend with it, confequently it 
has the ftrength of all the other rafters tideways added to it. 
To give an idea of the beft mode of framing roofs, and 
trailing up the timbers, we have delineated the feveral 
figures in the annexed Plate, of which the following is an 
explanation. Fig. 1, is a defign for a double roof, and is 
recommended to be ufed inftead of an M roof, upon all 
occalions where the fpan is great, and the pitch not very 
high. Fig. 2, is a defign for a Angle roof, of confiderable 
fpan; and the connexion of the roof in the walls is like¬ 
wise 
