268 THE STRENGTH AND ELASTICITY OF IRONBARK TIMBER 
whole structure in the short space of time allowed for its orectita : 
other kinds of timber were used for the piers, ne as stringybark, — 
ash, messmate, apple, box, spotted and white gu which a large ~ 
proportion had to be cut when the sap was up, Sielend of during © 
the winter months. In consequence of this dry rot soon appeared — 
‘in many of the piles, and this decay, together with the cavity or 
pipe in some of them, reduced considerably their sectional area. — 
There is considerable uncertainty in the time which a timber ~ 
viaduct may be supposed to last in this colony. i 
The following experiments have been made in a to compare a 
the relative merits of two forms of timber viaducts as used in New 
South Wales and Victoria. In New South Wales fort spans of 29 ft. 
6 in. and 26 ft. respectively, four compound beams are used on piers 
formed with round piles which are ron Seen aces see plate 12. 
In Victoria and in Tasmania for span a strutted 
timber beam is used on timber trestle sae see sieti e 14. Two 
model compound beams were constructed em one-eighth the 
size of those used in the actual viaduct, see fig. Two mod 
account of the difficulty in making small scale er the 
strength however cannot be much affected by the alteration 
The ate of timber used in model compound beam are as— 
follows 
© baie sie . 444in. x yin. x Ib inw = 
ead Aicnioines te RO oe OE 5 
4 half-corbels... 74,, x 14, x14, = 
199 cubic inches 
45 ” 
34 ” 
Total quantity of timber = 278 
The quantities used in strutted beam are :— 
Zheams... ... 444in. x 2 in. x Ofin. = 132°75 
ts... be i ee eh eS z 
l straining piece 26 ,, x 2 , x 14,, = 66°00 
Total quantity of timber 254-00 ; 
e ironwork in the models is about the same in each, although 
in the Victorian viaducts there is considerably more ironwork than 
in the New South Wales viaducts. 
models were tested horizontally in the machine, the load 
end of projecting nen ene (representing effect of weights of 
adjacent spans) w s applied by means of a hanging wel 
attached to a eines passing over a pulley. This bine amen is deno 
by w, in the following tables of experiments, while the weight 
applied i in the centre of model is denoted by W. 
