Function of feature on a semaphore shunt signal (NSW, Australia)

[ElStoko]

[/URL][/IMG]

Here's an image of mechanically-operated, lower-quadrant, triple-semaphore shunting signal in use on NSWGR in the 1950s. At the bottom you can see 3 face-plate levers with lozenge-shaped balance weights. The non-weighted side of each lever is pulled down by the action of a signal lever in a remote signal box via a length of signal wire. This action lifts the weighted side of a lever upwards, and, in doing so, forces upwards a vertical rod connected to the weighted-side of the lever. You can clearly see 2 of these rods rising above the balance weights in the image. In the technical drawings I have of typical lower-quadrant semaphore signals in use by NSWGR, the vertical rods attach the spectacle side (right) of the signal arm's pivot point, and so an upwards motion of the rod causes the semaphore to lower and the spectacles to rise (as it should for a lower-quadrant signal). However, if you inspect where the vertical rod attaches to the semaphores in the image, you can clearly see that the attachment points are to the semaphore side (left) of the signal arm's pivot point. To create a downward motion of the semaphore, the action of the vertical rod would have to be mediated through a second lever. These second lever points are (I think) indicated at the points where I have drawn black arrows in the image.
My queries to those with the knowledge are:
a) Do you agree that a second lever was employed?
b) If so, why are the second-lever pivot points occupied by box-shaped structures protruding from the signal post? These appear too large to be simply providing a body to which pivot pins are attached.

 

[ElStoko]

Could the "box-shaped structures protruding from the signal post" at the second-lever pivot points be signal operation detection relays?
Point-operation detection circuits were in use on NSWGR to confirm that the switchblades of critical junctions did complete a changeover of switch direction. I don't recall coming across an example where a detection circuit was employed for the operation of semaphore signals. Normally a signal could be sighted from a signal box, so its operation could be confirmed by day by observing the semaphore position, and by night by sighting the front or rear lights. At Hawkesbury River, where the image of the signal was taken, the signal box did not have a good view of the yard. The signalman could not sight this shunting signal without leaving the signal box and walking 50 metres. The introduction of second small lever in the action to control each semaphore might have been motivated by the utilisation of the lever action to open and close a relay switch which could break and make an electrical circuit that causes a light to extinguish or shine in the signal box (under or above the lever that operates the semaphore).

Anyone out there - perhaps an ex-NSW signalman - confirm that such signal operation detection circuits were employed, and that a second lever as described was used to implement such circuits?

 

[ElStoko]

With the aid of some further images, I've decided that the semaphore-operation rod linkages and the those of the semaphore-operation-detection rod linkages are independent of one another. What appeared to be a second lever in the semaphore-operation linkages is a link in the independent semaphore-operation-detection linkages. A detection link attaches to the semaphore side of the semaphore pivot point and (hidden from view in the image of post #1) an operation link attaches to the spectacle side of the pivot point.
The additional images of 3-arm shunt signals at Hawkesbury River are on pages 25 and 27 (upper image) of 'Visiting Uncle Fred', Adrian Rynberk, Australian Railway History, September 2013. The image in the first post is extracted from page 6 of 'Hawkesbury River to Woy Woy', Ian Wallace, Byways of Steam #10.