WOODEN HOUSES—THE FRAME
Almost any variety of wood will suffice for the frame of the house, provided it does not twist and spring out of shape too much before or after it is put into the building. Since the sills are to be placed on solid, continuous walls, they need not be large. The only objection to box and small sills is that they may allow too easy access of air and rodents from the walls of the rooms to the cellar, and vice versa, unless the spaces above the sills and between the studding are bricked in as high as the top of the first tier of joists. A rough floor laid before the upright studding is placed is shown in [Fig. 54]. This first floor should be laid diagonally, for the one which is laid immediately upon it should not be placed either parallel or at right angles to the boards of the first floor, or parallel with the joists. A little reflection will reveal the reasons for all this.
Fig. 54. The rough floor laid before the studding is erected.
Joists should be bridged. [Fig. 55] shows the more common method of bridging. The joists may be 2 × 8 in small, inexpensive houses, and 2 × 10 or 2 × 12 in large ones, bridged once in a 12-foot span, twice in a 16-, and three times in an 18- or 20-foot span. The bridging is of the utmost importance and should never be omitted, as it serves to strengthen the floor joints and prevents the disagreeable trembling of the floors so annoying in many of the older houses.
Fig. 55. Bridging the joists.
The studding for a balloon frame is either 2 × 4, 2 × 5 or 2 × 6, and the length desired. The 2 × 4 studding are too light for an ample two-story house, and they do not give enough thickness of wall for the most desirable window- and door-jambs. The doors are not held firmly in place, and when they are closed quickly by the wind or by children, the plastering is injured. Studding 5 inches broad, fortified by outside diagonal boarding ([Fig. 56]), gives the ideal conditions unless the house is unusually large, in which case the studding should be 6 inches broad. The diagonal boarding costs a trifle more in material and labor than the horizontal, but it is so much superior that the extra expense may well be incurred. Every board forms a double brace, one where nailed to the studding and one where the siding or “clap boards” are nailed to the rough boards and the studs. Nothing has yet been discovered which is so satisfactory, and which gives such strength and protection to the frame as does this preliminary diagonal boarding, covered with paper. When completed it forms a wall open enough to prevent dry rot and tight enough to prevent the entrance of wind.
Fig. 56. A wall strengthened by diagonal sheathing.
The second-story joists rest on stringers or light girders 1 × 5 inches, as shown in [Fig. 57]. If the girder is set flush with the inside of the stud, A, the laths must lie directly upon the face of the girt. This gives no room for the mortar to form clinches behind the lath. This 5-inch girder swells when the mortar is put on and shrinks when it dries, which may result in a crack in the wall in the angle near A. Since, by reason of faulty construction, there are no clinches behind the lath, the plastering becomes loosened, and this is likely to be the beginning of serious trouble. If the girder is let in so that its face is not flush with the inside of the stud and then furrowed out with small pieces of lath, the effects of the shrinking of the girder will be obviated and room will be left for clinches behind the lath.
Fig. 57. Second-story joist.
In windy, cold climates, where lumber is at all abundant, a second boarding may be placed inside, covered with paper and furrowed out with a single thickness of lath to allow, as in the former case, the formation of clinches. There is no objection to boarding horizontally on the inside, if the outside has been boarded diagonally. The term “rough boarding” has been used, but it should be said that the boarding which forms the first covering, sometimes called sheathing, should be brought to uniform thickness and matched or rabbeted.
Wherever greater strength of wall is desired than can be formed by a single 2 × 5 studding, as at the corners, or by a single 2 × 10 joist, as where partitions are to be placed, it is better to spike two or more pieces together than to have pieces sawed of the dimensions desired. These made-up pieces or timbers are stronger than solid pieces of the same character and dimensions, since the continuity of the cross-grain of the wood is broken in the made-up pieces. In the construction of large bridges the timbers, where exposed to the weather, are made up of smaller timbers, since they are then not only stronger but more durable and less subject to dry rot than if they are solid ([Fig. 58]).
Fig. 58. Construction of a large bridge.
Plates are made up of material 2 inches thick and as broad as the studding is wide, doubled, with joints mismatched. This most valuable principle of building up timbers of several thin pieces is a somewhat recent practice. Where very large timbers are required, as in trussed or self-supporting roofs, the timbers of which are not exposed to view, they are frequently made up of boards 1 inch thick and as broad as the vertical dimensions desired. This method is sometimes used in constructing timbers for both houses and barns ([Fig. 59]).
Roofs of houses are, of necessity, extremely variable, as the house is not planned to suit the roof, but the roof to suit the house. Flat metal roofs of all kinds should be avoided, as far as possible, on the farm house, however well they may be adapted to buildings in the city. Metal roofs are not objectionable in themselves, but only when they are laid flat on farm houses.
Fig. 59. A made-up plate, constructed of boards.
Fig. 60. Showing the principle of construction of [Fig. 59].
The pitch of roofs, like their shape, is also variable. Nothing below one-third pitch should be used except for special conditions. In [Fig. 38], page 127, is an illustration of the common pitch of roofs in fashion fifty years ago. Some roofs were even flatter than the one shown. The fashion now is to construct house roofs with nearly or quite half pitch. While steep roofs are desirable if made of wood, there is some danger that the change from the nearly flat roof to the steep one will be carried too far (see [Fig. 13], page 95). Various pitches of roofs are shown in [Fig. 61]. Steep roofs do not require as strong rafters, thrust less upon the plates, are more durable, and are less likely to leak than flat roofs.
Fig. 61. Pitches of roofs.—¹⁄₂, ¹⁄₃, ¹⁄₄, ¹⁄₈.
Since roofs are of various pitches, they require rafters of various lengths and bevels. Farmers and many carpenters have much difficulty in getting the length and bevels of both rafters and braces. Most carpenters’ squares have so-called brace rules stamped upon their tongues.[3] These give the length of the brace for the shorter and more common runs,[4] but they do not give the angles of the ends of the brace. Then, too, the length is given in inches and hundredths of inches, and carpenters’ squares are not divided into hundredths, so this complicated brace-rule is as useful as a steam whistle on an ox-cart.
[3] The short end of the square.
[4] The perpendicular and horizontal distances covered by the brace.
The methods by which the length and bevels of any member of a frame which departs from any other member at an angle are so easily understood that the wonder is that all are not familiar with them. For a simple illustration, let it be supposed that rafters for a building 18 feet broad, with one-third pitch, are to be laid out ([Fig. 62]). The rafter, R, takes the form of a brace. The run is 9 feet horizontally or half the width of the building, and 6 feet perpendicularly. If the square be laid upon the stick designed for the rafter, as 6 is to 9, one side of the square will give the shorter and the other the longer angle or bevel ([Fig. 63]). If the square is laid on 12 times at 9 and 6 inches, it will give the length of the rafter, for 12 times 9 is 108, half the width of the building, and 12 times 6 is 72, the height of the peak above the plates. If the square is laid on 18 × 12 inches, the proportion is preserved, and hence the angles; the square would only have to be laid on six times.
Fig. 62. Laying out a roof.
Fig. 63. Laying out a rafter.
Fig. 64. Laying out a timber.
Fig. 65. A brace.
Consider a building 20 feet broad and 6 inches above one-third pitch. The half of 20 feet equals 10 feet, or 120 inches. Seven feet 2 inches (86 inches) is the height of the peak above the plate. It is quickly seen that this problem, like the other, can be solved in more than one way. If the long end of the square is laid on at 20 inches and the short end at 14¹⁄₃ inches, and this is repeated six times, both the bevels and the length will be secured ([Fig. 64]), for 6 multiplied by 20 equals 120 inches, half the width of the building, and 6 multiplied by 14¹⁄₃ equals 86 inches, the height of the peak. Or the long end of the square might be laid on at 24 and the short end at 15¹⁄₅ five times, but squares are not marked in fifths of inches, hence the previous method would be best.[5] The same results would be reached by laying the square on at 15 and 10³⁄₄ inches; eight steps would then be required instead of six. The longer and fewer the steps within the limits of the square, the better.
[5] Since the square is laid on, see [Figs. 61], [62], in the same manner as for cutting a stair; each one of these spaces is called a “step.”
If it is desired to cut a brace 3 × 4 feet run, 3 steps, using the lengths 12 and 16, will give both the length of the brace and the bevels ([Fig. 65]). Take a rafter which has a projection requiring a notch to be cut in the lower side, and the same rule will apply. The line A, [Fig. 66], is horizontal and the face of the plate is perpendicular; therefore, the line B must be at right angles to A. The only thing now to be determined is how deep the notch shall be, for it is evident that if the line A represents the long end of the square and B the short end of the square, the notch will fit the plate.
Fig. 66. Adjusting to the plate.
Fig. 67. The rafter.
Fig. 68. The rafter trimmed on the outer end.
That part of the rafter which extends over the building may be reduced in size, but usually it is well to leave it entire (as in [Fig. 67]) if the house is large. If the lower end of the rafter should appear too heavy, it may be treated as in [Fig. 68]. The bevels at the ends of the rafters are the same as at A and B ([Fig. 66]).
The outlines of a story-and-a-half house, which form is most undesirable for various reasons, are shown in [Fig. 69]. The chambers cannot be well lighted or aired. The outlines of the room interfere with the placing of furniture, and such chambers are far more uncomfortable in warm weather than are those in two-story houses. It will be seen that the collar-beam, C, must be placed so far above the foot of the rafters in order to get a fair height of ceiling, that it has little binding power, and that the building cannot be tied together at the plates in the center, since the tie would interfere with the door in the cross wall. It will also be seen that the second-story joists are so far below the plates that their power to hold the building together is small. Many of the one-and-a-half-story houses have “sway-backed” peaks because of this faulty construction. (See [Fig. 35], page 124, broken-back house.) If story-and-a-half houses must be built, then they should be covered by roofs having at least one-half pitch, in which case the collar-beams could be placed relatively lower and the thrust on the plates would be very much diminished by the steeper roof ([Fig. 70]). One-, two-, three- or more storied houses are easily and certainly prevented from spreading since one tier of joists always coincides with the foot of the rafters, to which they can be securely fastened. Fortunately, the story-and-a-half house is less constructed than formerly.
Fig. 69. Outline of a story-and-a-half house.
Fig. 70. Half pitch and an efficient collar-beam.
CHAPTER IX
BUILDING THE HOUSE, CONCLUDED.—OUTSIDE COVERING, PAINTING
That part of the house which of necessity must be exposed to the buffetings of snow and rain, wind and sun, should be considered more carefully than any other part except the foundation. If economy demands, the doors, floors, bath rooms, and wardrobes may be of plain and inexpensive material, for later they may be replaced when means justify additional expenditure; but if the outside covering be faulty, the house is a partial failure from the beginning.
The first principle to be observed is to place all projections intended to serve as water-tables at somewhat acute angles, for if placed at nearly right angles with the sides of the house, rains accompanied by heavy winds will certainly reach the framework. The water-tables which crown the top of the base-board are more exposed than those which are higher up, and therefore should be steep and rabbeted to prevent the water from reaching the sills. The too usual method is shown in [Fig. 71]. An enlarged view of a better style of water-table is shown in [Fig. 72].
Fig. 71. A faulty water-table.
Fig. 72. A good water-table.
Outside window frame sills which have insufficient pitch tend to become water-soaked, and not infrequently the lower member of the window itself rots by reason of the water which drives in and remains under the sill of the window for considerable periods of time. [Figs. 73] and [74] show perfect and faulty methods of construction.
Fig. 73.
Perfect construction of window sill.
Fig. 74.
Faulty construction.
The siding of a house for various reasons would better be put on horizontally, although material put on this way, unless it is kept well painted, is not so durable as when placed vertically. The horizontal covering is more beautiful, lends itself better to the numerous openings, and gives better protection from cold and wind than does the vertical covering. If the building is not to be painted, then the covering would better be placed vertically. Nearly all wooden houses are covered with either thin lap-siding or inch siding, prepared in various ways and known by various names. The inch or novelty siding was first introduced in the West, and costs but little more than the lap-siding, because, being thicker, it can be made of somewhat inferior lumber. The novelty or rabbeted covering gives greater strength to the building and is much more quickly and cheaply put on. It may be said that this style of covering is extremely faulty if placed on the building in the usual way, namely, before the doors and windows and corner boards are in position. If the same method of placing the material be practiced as in placing the lap-siding, then the objections to this class of siding disappear to a certain extent. The diagram, [Fig. 75], shows the novelty, or drop, or O G siding (A), the rabbeted (B) and lap-siding (C). It will readily be seen that if a drop (A) or rabbeted (B) siding be put on before the window frames are placed, as is the usual custom, an opening (x) is left under the facing of the window frame which extends through to the studding. This permits the rain, in a driving storm, to pass horizontally along this opening to the studding and then downward along the framework of the building. Many instances could be cited in which these openings have had to be filled by triangular blocks of wood or putty, and even then the water was not entirely excluded.
Fig. 75. Forms of siding.
This method of covering houses or even barns with this new kind of siding is usually disappointing and wasteful of material. All that is gained is a little more facility and cheapness in placing the covering. If it is put on, as it should be, after the window and door frames are set, it is more difficult and more expensive to place than lap-siding.
No way of covering a wooden house has been found superior to the one-half inch lap-siding with joints tight enough at the frames and corners, in conjunction with the paint, to make water-tight joints. The lap should not be less than one inch, and the nails should be so placed that in case of considerable shrinkage in the siding the inside will give or even check, instead of the outside (z). If made as at y, the outside will check. This implies that the nails are to be driven rather more than one-half inch above the edge of the siding. The nails which hold the outer covering should either be set and puttied, or the heads should be left even with or slightly above the surface of the wood, that the paint may cover all parts of the nail head. If the nails are driven too far in the heads are not fully covered and protected by the paint, in which case they will rust and present an unsightly appearance.
Some one has said that if a woman’s feet, hands, and head are well and appropriately clothed, the balance of the dress may be plain and simple, and yet she will have an elegant appearance. So, if a house has a good foundation and a suitable and well-placed roof, the balance of the outside may be extremely plain and yet it will be beautiful. Some of our modern houses rest on unpointed, poorly constructed, and narrow foundations, are bedecked with peaks, pigeon lofts, and dog-eared cornices, and remind one of the suspenderless, barefooted darky crowned with a cast-off silk hat.
If the foundation is too small and shabbily built, no amount of paint and cornice can relieve the house from a look of shabby gentility. A few brown or cream-colored stones or bricks, when placed on the outside of the foundation where it shows above ground, will give dignity, beauty and a substantial look to the whole house. It may do for it what a nickel does for one’s shoes.
The roof of the farm house, and for that matter of all other houses, should, in the trying climate of America, have an ample projection. An abbreviated cornice may be admissible if the building is constructed of stone which is of sufficient density to resist the American tooth of time. [Fig. 76] shows a section of an abbreviated and a well extended cornice. The house which has this short-cut cornice stands within a few hundred feet of the one with the wide projecting eaves. During the past twenty years it has been necessary to paint the former twice as often as the latter.
Fig. 76. Deep and narrow cornices.
The roof covering would better be of slate or tiles, for the time has passed for building temporary, make-shift houses, though they might have served their purpose well in a new and rapidly developing country. With rare exceptions, the houses to be built in the future should be permanently located, well built, and of durable material. The slates which compose a roof should be not more than 8 inches wide and should not be put on roofs of less than one-third pitch, since they are only double-lapped and do not lie as closely, one upon the other, as do shingles, which are laid triple-lapped. Slate and tile roofs are comparatively heavy, and hence require stronger roof structures than shingles.
The roof boarding for slate roofs should be matched—tongued and grooved—and covered with paper to prevent cold and draughts of air from passing into the attic. Since slates, on account of their somewhat rough surfaces, do not lie closely together, the wind is likely to pass through the cracks in the roof, if there are any, and carry snow and rain into the upper part of the house; therefore the roof covering immediately under the slates should be virtually air-tight. The roof boards for a shingle roof should be narrow and laid with openings of from 1¹⁄₂ to 2 inches between the boards. Rain and snow seldom drive up and through the shingle roof, and since wooden roofs are more likely to rot out than to wear out, the more perfectly the shingles are dried out after a storm the better. The narrow roof boards and the spaces between them allow the shingles to dry quickly, and therefore are better than matched boards.
The short, or common, shingle of commerce is 16 inches long, ³⁄₈- to ¹⁄₂-inch thick at one end, and ¹⁄₈ of an inch at the other, and is computed at 4 inches wide. A bunch of shingles contains one fourth of a thousand. It should have 25 double courses and the band should be 20 inches long. Not infrequently there is a course or two wanting, or the bands are an inch or so short. Having this data, one can easily determine if the bunch is of legal size. A little cheating is not uncommonly done by placing the shingles in the bunch loosely. This can be detected by examining the bunches at the thick ends of the shingles.
Theoretically, 1,000 shingles should cover 10 feet square, or 100 square feet, known in carpentry as “a square,” if the shingles are laid 4 inches to the weather. Since shingles are usually laid 4¹⁄₂ to 5 inches to the weather, 1,000 shingles should cover about 120 square feet. Two-thirds of the lower part of the roof may be laid 4¹⁄₂ inches, and the upper third 4³⁄₄ or 5 inches to the weather, if the roof is not flat.
If shingles are treated with lime water or diluted gas tar, or be painted as they are laid, the life of the roof may be prolonged. The painting of roofs with tar or common earth or mineral paints, after they are laid, does little or no good in preserving them. Sometimes painting is resorted to to make the roof harmonize with the color of the sides of the building.
Neither extremely narrow nor extremely wide shingles are desirable. Those from 3 to 6 inches wide, when carefully laid, are satisfactory. Each shingle should receive but two nails; one is usually enough, and these should be placed about ³⁄₄ of an inch from the edges, and about 1 inch above the point where the butts of the next course will come. When the courses above are laid upon the shingle having but one nail, two or three other nails, which are driven in the courses above, will serve to help hold it in position. The joints of shingle roofs should be double broken: that is, the joints in the shingles of one course should not coincide with the joints of the first or second course below. Consult [Fig. 77].
Fig. 77. The laying of shingles.
If two nails be driven in the sides of an unseasoned shingle, when it shrinks it is likely to split in the middle; and in laying a roof the joint immediately above the course under consideration is likely to come at or near the middle of the shingle, which splits by reason of the shrinking. The case is still worse when three nails are put in a shingle, for then it is almost certain to split in the middle and immediately in line with the joint in the course above.
Unscientific placing of shingles and insufficient mixing of mortar results in an unsatisfactory house, both inside and outside, however good the materials may be.
Fig. 78. A veneered wall.