I'm reviving this thread as, after years of searching, I'm now the proud owner of a Cambridge Dolezalek Quadrant Electrometer.
My example is a Cambridge Scientific Instruments Company Ltd model, dating from around 1906. Unlike the OP's it has it's inputs brought out on the underside and has a square window. I was also lucky that it came complete with mirror assembly and very luckily, when I opened up the quadrant, I found the original silvered paper vane which had become detached.
The company history is quite interesting, one of the founders was Charles Darwin's son. It was founded in the late 19th century, later getting into electrical measurement instruments, and introducing the quadrant electrometer in 1906. In 1919, the company merged (took over) with the Robert W Paul Instrument company to become the Cambridge and Paul Instrument Company, and then in 1924 changed it's name to the Cambridge Instruments Company, adding the familiar 'bridge' logo seen on the OP's example. Robert W Paul was an interesting character, in addition to his Scientific instruments company, founded in 1891 (I have one of his early High Sensitivity suspended coil Galvanometers, and also one of his Unipiviot ones), he was also the pioneer of cinematography in England, demonstrating moving picture projection in 1896 (around the same time as the Lumiere brothers in France) but after Edison. He had his own film studio.
I've attached some photos of the repaired Electrometer. I've re-fitted the vane, not easy because it is extremely fragile and had been torn from the mirror assembly. Whilst I have replaced the suspension, with a fine piece of glass fiber from a fibre-optic bundle (instead of the piece of cotton that saved the mirror from being lost), this still comes in at around 40um, which is still well short of the 4, 6, or 8um fused quartz suspensions that it originally came with. I've no idea where I would find one of those - I'm sure I have some single mode fibre somewhere, which has a core diameter of 7um which might get me close.
With the correct 6um suspension, these electrometers were capable of measuring below 0.1V potential differences across the quadrant inputs once the vane was charged to around 100V - it has a true differential floating inputs. I had always associated 'passive' electrometers with higher voltage measurements.
I've attached some photos, together with the relevant 1906 pamphlet - it's interesting to see that, at the time, their phone number was 'Cambridge 6'! Hopefully this is of interest to the OP and maybe others.