ON ARCHITECTURE AND GARPENTRT. - 1^7 



on each other, resist the force of a weight compressing them, as effectually as if 

 they formed hat one piece: biit they have no sensible cohesion to enable them 

 to withstand a force tending to separate them, and if they are required to co- 

 operate by tlieir cohesive strength, some mode of uniting them must be found. 

 For this purpose, it is generally necessary to sacriiice a considerable portion 

 of the strength of the materials employed. The most usual mode is to place 

 the ends of the pieces side by side, first reducing their dimensions, where a 

 regular outline is required ; and to jjrocure a firm adhesion between them by 

 means of external pressure, or to employ the natural adhesion of some parts 

 which are made to project beyond the rest in each piece, and receive in their 

 interstices the corresponding projections of the other piece. 



Where the adhesion is produced by external pressure only, it is of advan- 

 tage to subdivide the joints into a considerable number of parts, as is usually 

 done in the masts of ships, and to make the junction of any two pieces, fol- 

 lowing each other in the same line, as distant as possible from any other 

 junction; for in this manner, the loss of strength may be diminished almost 

 without limit, provided that the distance between the joints be great enough 

 to afford a firm adhesion to each part. The junction may also be formed by 

 an oblique line ; but the obliquity must be so great that any lateral pressure 

 may increase the stability of the wedge, the length being in a greater pro- 

 portion to the depth than the pressure to the adhesion that it occasions; and 

 the pieces must be pressed together very forcibly by means of hoops or bolts. 

 (Plate XIII. Fig. 166 . . 168.) 



Where the natural adhesion of some projecting parts in each piece is em- 

 ployed, the projections must be sufficiently long to secure their strength,, 

 and they must be as Httle. prominent as possible, partly because the contigu- 

 ous piece nuist be excavated for their reception, and partly because their 

 strength is diminished when they project more than one sixth of their length. 

 A beam united to another in this manner is said to be scarfed. (Plate XIII. 

 Fig. 169.) 



In order to preserve the strength of a compound beam, intended to resist a 

 transverse action in a particular direction, it is necessary to avoid, as much 

 as possible, reducing the depth of the beam in that direction, and to secure 



