121 
sion is not recommended, as no degree of elasticity in the filling sub- 
stance can increase the resistance to deflection. The tamping of the 
ballast under steel ties must be SO done that the middle of the tie is 
either empty or at least not compactly filled with ballast, while lor a 
distance of 12 to 15 inches on each side of the rail the ballast must be 
well tamped. 
It is also import. int that during tin 4 first lew months after the ties 
are laid tbe road should have great and constant attention until the 
bed becomes consolidated. The cost of this ought to be considered as 
part of the cost of laying' the road. 
The metal tie acts best on firm ground. Thus on the Liege-Liui- 
boarg line there were 20,000 metal ties which had not beeu touched 
for twenty-two months and the road continued in a most satisfactory 
condition, working trains running at a maximum speed of 38 miles per 
hour. Lengths of ballast of medium quality in which wooden ties ap- 
peared to be dry have proved unsatisfactory with iron ties. By means 
of their pumping action the latter draw the wet up from below and 
work the ballast into mud, making a solid bed impossible. This work- 
ing up into mud occurs also with wooden ties, but only after the bal- 
last has become completely impermeable for water, and requires 
renewing anyhow. 
The reasons for this difference may be stated as follows : 
(1 ) The iron ties have twice the deflection of the wooden ones. 
(2) On account of the accurate fasteuing of the rail to the iron tie, 
it shares the whole vertical motion of the rail, whereas with wooden 
ties the play of the foot of the rail in the dogs and the compressibility 
of the timber both tend to lessen the deflection of the tie. 
(3) The hollow body of the iron tie is very favorable to the formation 
of an air-tight cavity, inducing the pumping action above referred to. 
(4) The under surface of the wooden tie lies twice as deep as that of 
the iron one. 
The working Up of the ballast into mud by iron ties, on certain trial 
lengths, has interfered with their more general introduction, whereas 
the failure should have been put down to the inferiority of the ballast 
and its impermeability to water. It would be found that in similar 
ballast an accumulation of moisture at the bottom of wooden ties takes 
place. 
Good ballast must fulfill the following two conditions : 
(1) It must be capable of being beaten up under the tie into a firm 
mass so as to afford the greatest possible resistance to the deflection 
when the load comes upon the tie and the greatest possible resistance 
to being shaken loose. 
(2) At the same time the material must be such that between the ties 
there may be in it interstices for the passage of water, but also the 
greatest possible cohesion and friction, to prevent slipping and to dis- 
tribute equally the pressure of the load. 
