58 
opinion that in the long run its adoption would be found to 
be beneficial to the shareholders of our railways, and it 
would contribute in some degree to the safety of the travel- 
ling public. 
There are two other sources of accidents on our railways 
that require notice — one, the system, now so prevalent, of 
centralising the signals, and the other, the breaking and 
making up of trains on the main line. 
The centralising of the signal handles into one box may 
possibly possess some advantages in saving wages, and also 
in placing the signals and points connected with or depend- 
ent upon each other within the control of one man ; but 
may not this be carried too far ? When the centralising of 
signals requires the man in charge to have his attention 
directed to two different trains -at the same time, and per- 
haps coming in opposite directions, it does create contingen- 
cies of a nature that will, at some time or other, lead to 
accident, and it is unreasonable to expect a man at such 
critical junctures always to do the right thing. Another 
objection to the centralising of signals arises from the work- 
ing of the distant ones. The mechanism required to form 
the connection between the signal box and the signal itself, 
is on account of the distance, liable to derangement, being 
affected by frost, heat, and rain, and repairs and adjustments 
are frequently necessary, thus creating another class of con- 
tingencies that may lead to accident. And it may be fur- 
ther observed that it does sometimes occur that the distant 
signals are beyond the observation of the signalman in his 
box, and is always so in thick weather ; so that he has no 
chance of knowing, in such cases, whether the signals 
answer to his workings in the box or not. 
The breaking and making up of trains on the main line 
has been the occasion of many accidents, and its continu- 
ance, especially upon lines having a large traffic, must lead 
to similar results. It needs no argument to show that a 
