First did you try reversing the leads to the track? The whistles were solid state and would only respond to a positive dc offset.
Second, this may be a very early circuit designed by a former IBM engineer and past TCA president Al Ruocchio. He showed me his circuit about 20 years ago at York. It does work but requires more than the usual 1.5Vdc offset, actually more like 10Vdc offset. Not sure if this is what you have.
Try reversing the wires first. Not sure how to solve the second problem without building his special activation circuit.
Lou N