642 PROCEEDINGS OF SECTIOX H. 



DiflEerent combinations of weights at A, C, and E are experi- 

 mented with, and each combination is tested four times, one result 

 being obtained with the beam in one position, a second with the- 

 beam turned upside down, a third with the beam turned end for 

 end, and a fourth with it again turned upside down. This process 

 eliminates the effects of any want of uniformity in the beam. 



We average tlie results obtained in this way. 



Next, in elivi'nitite the effect of the beam's men weight (see Fig. 

 29). Ii the beam were weightless, it would be level when supported 

 at B, C, and D, before any w^eights are ap})lied. Not being weight- 

 less, it assumes a form, resembling that shown in an exaggerated 

 Avay in Fig. 29. Certain forces would 'be required at the ends and 

 middle to bring these points level with the tops of the posts. The 

 beam will not even then be quite level, because there will be sagging 

 at intermediate points ; but if, by means of spring balances, forces, 

 A\ B\ C"\ D\ and E^ be measured, and compounded with the 



fcirces Ri, R2 cos ^ i^2y "^c., previously found, we obtain a result,. 



approximately free from the effects of the beam's own weight. This 

 comj)ounding is to be pei-fonned as follows, viz. : — 



(1) Subtract C^ from the reading obtained with the spring 

 balance at mid-span, and thus obtain a corrected value for Ro- 



(2) Subtract B^ and D^ from the readings of the spring- 



n 



balances at B and D, and thus get a corrected value for R2 cos ^• 



(3) Add A^ and E^ to the weight applied at the ends, and thus 

 obtain corrected values for Kj and B3. 



A^, B\ &c., ai'e taken to be the averages of the values obtained 



with the bar in foui' different positions, these positions being obtained 



l)y turning the bar upside down and reversing, end for end. 



B B 



Strictly, the forces (R., — C), (R„ cos -— BMaud (R,, eos-_DiY 

 ' ~ "2, '2. 



e . 



sliould iDe in the pro})ortioii 00s « • 1 • cos ;,; and, if the results vary 



mucli from tliis, we may have to find, by repeated trials, values for 

 the weiglits and forces, measured by spring Ijalances, which will 

 ultimately satisfy this condition. 



COAX'IADIXO ]vE.\lAUKS. 



At the commencement of this pu])er, it was said that the use 

 of trussed beams was likely to increase. This opinion was based 

 upon the cheapness of structures of this iyj)Q as compare<i with 

 short braced composite girders, or with beam bridges whose piers 

 are close together. A completely braced girder nu\y contain less 

 timber and steel than a trussed beam of the same span ; but the 

 labour in constructing the girder is very much greater than that 

 required to construct the trussed beam, and the material cut to- 

 waste when shaping the smaller members of the braced girder might 

 go far towards making up for the gi-eater amount of material put 

 into tlie trussed beam. Compared with a beam bridge, on pile piers, 

 when the weight of the piles and the labour in driving them are 

 considered, the advantage will frequently lie with the trussed beam.. 



