However, the line never progressed past the Belmont (now Ascot) racecourse: the local parliament realised that the route was impractical and the present South West Railway following a more direct (and easier to describe!) route was substituted in lieu. In 1886 the branch was built as far as the Swan River, however, it stopped there for some 11 years until October 1897, when the last 60 chains was constructed bringing the now duplicated line across the river and into the ultimate terminus.
Initially, the line worked under the rules applicable to staff and ticket working. By March 1897 absolute block working (using Winters Instruments) had been introduced to the Branch, although the exact date appears to have gone unrecorded. Precisely what form of absolute block working applied when Staff and Ticket working was abolished is not entirely clear: early references to race train traffic on the branch refer to the cabin at the Racecource (this must be a reference to a "traffic office“ and not a cabin in the strict sense) being opened as a block telephone cabin. It is possible that Winters Instruments were initially not used when the line was duplicated: this is a subject for further research. By 1901, absolute block working appears to have been in use only when the line was used for race trains. The General Appendix of the same year provided that when Belmont Racecourse station (or Racecourse as it was then known) was unattended (which was normally the case) only one train was allowed on the branch at any time.
Unlike electric staff machines, which could be operated either automatically (both ends of the section being unattended) or semi-automatically (with only one end of the section being attended), Winters Instruments required manned boxes at each end of the section. The same was true for telephone block working to have any benefits: a guard could attend to giving line clear once his train had arrived but this would be a slow process, defeating the object which was to increase track capacity. In any event, from the time of the 1901 General Appendix, Bayswater signalmen were enjoined to allow only one train on the branch at any one time when racecourse was unattended, trains proceeding on a “caution”. Presumably the signalman at Bayswater clipped the lever of the branch Starting signal to remind himself not to dispatch a following train until the preceding one had returned to his station. This restriction probably did not create any operating headaches, as normal traffic on the branch was restricted to a handful of trains a day, with the line only really coming to life (and a very busy life at that) on race days.
The first reference to River Bridge having any safeworking function was that in Weekly Notice No 22 of 1898, when the notice announced that on 1st of June 1898 block Instruments were to be connected at Falkirk (todays Maylands) and Racecourse River Bridge. The first reference to a cabin at River Bridge appears to have been in Weekly Notice No 45 of 1898, which announced that a block post at River Bridge was to be opened on 9th of November 1898. In any event, its existence was considered important enough to be accorded an entry in the 1901 General Appendix, which is worth repeating in full:
It might appear rather difficult to operate a signal box with no fixed signals. However, it is quite probable that the procedure applicable was similar to that as applied to Fenians’ (the only other block post lacking fixed signals) as the following entry from Weekly Notice No 43 of 1903, detailing the procedures there reveals:
“....Four flagmen will be provided at Fenians' Crossing and 17 Mile cabin, to act respectively as Up home, Up distant, Down home, and Down distant Signals. In the absence of any instructions from the signalman, a red flag must be maintained and exhibited by the flagmen to the driver of an approaching train. As far as practicable the flagmen must shield the flag (by holding it in front of their bodies) so that it will not be seen by an engine-driver coming in the opposite direction. The flagman acting in place of a Home signal will receive all signals from the cabin, and act accordingly. The flagman acting as a distant signal must be in vision of the flagman acting as the Home signal and must repeat the signal exhibited at the latter place.
Should a train approach while the danger signal is exhibited at the home, the danger signal must also be exhibited at the distant, but if the road is clear between the Distant and the Home, the flagman will, on the near approach of the train, and after it has been brought quite or nearly to a stand, lower the red flag and announce to the engine-driver “Distant signal;” the train may then proceed slowly to the flagman acting as the Home signal. Flagmen are directed to stand on the left hand side of the line, and to hold the flag with both hands, the lower hand holding the bottom, so that a large surface will be exhibited. The signalmen at the intermediate cabins must see that the flagmen under their control thoroughly understand their duties before they are told off to their respective positions. “
Whether or not extra tall flagmen were employed as Distant signals (so as to provide better sighting distances) is not recorded.
River bridge was located 5 miles 48 chains from Perth, (which places it on the Bayswater side of the bridge) although on which side of the line it was located is unknown to the author, although aerial photographs taken in 1948 suggest that the box may have been located on the down side. The odds are that the box was a standard Western Australian Railways weatherboard and gable roof design, although the possibility that the box was one of C. Y. O’Connor’s skillion roofed NZ inspired monstrosities cannot be discounted.
River Bridge may not have ever been provided with fixed signals, there is some evidence however, to suggest that it may have, for a short period at least, worked a crossover. A February 1903 Weekly Notice entry announced that as a consequence of the cross-over road between the Up and Down line having been removed, Walkerden’s Brick Siding, situated on the Down side of line, Bayswater end of River Bridge, could hence only be worked by Down trains, and any vehicles picked up may be propelled to Belmont Station. Whether the crossover was worked from the signal box is not certain: as the points and the siding could have been operated from a ground frame or rigid levers.
If the crossover was in fact worked from the box this would render River Bridge unique, as being the only box in Western Australia that operated points only not signals.
The abovementioned crossover and siding were not the only points to grace River Box over the years. The bridge was made from West Australian Jarrah, which was far cheaper than iron or steel, although nowhere as durable, and the Bridge required major repairs several times during its lifetime.
The first such work was in 1908, when it became necessary to repair the Up spans of the bridge. The Up road between Belmont and River bridge was closed and a set of balanced points (set in favour of the Up direction) connecting the Down to the Up was installed.
Block working remained between Bayswater and River Bridge, with pilot working between Belmont and River Bridge. The relevant weekly Notice provided that the pilot was to accompany the train the entire length of the branch, and was to hand signal the train at River Bridge, except on Saturdays when the box was attended. Similar arrangement were adopted in 1918 when the Up and down spans were renewed (one wonders if there was something defective in the timber used for the bridge) except that River Bridge was unattended, and Staff and ticket working between River Box and Belmont was introduced.
When ‘cut in’, River Bridge worked with 17 Mile Block (Signal) Box (Belmont Branch) (as it was known in the early General Appendices) or, more latterly (and sensibly!) ‘Whatley’ on the Up side and Belmont on the Down. Throughout its existence Belmont 's facility was described as a cabin or a box, however, prior to 1922 this appears to have been a loose use of language, the facility being more properly described as a traffic annex.
During the first decade of the Twentieth Century River Bridge was regularly opened as a block post to facilitate the movement of race traffic, however, the frequency began to decline after about 1909. The start of World War One saw a drastic reduction in racecourse traffic, which continued until the Armistice, and the box does not appear to have opened at all during the War.
The return to peace did not see a return to normality: after 1918 the Box saw little use, being opened only for the end of year race meetings.
1921 saw a new suburban time table, which listed River Cabin for the first (and last!) time, although no services were timed to stop there.