At least the size and construction are similar. Apparently Edison-Ford museum in Florida has this and a whopping 75 kW lamps on their collections, but the link didn’t work.

The fabric of the arm of the chair is plainly visible through the neck of the bulb. The fabric pattern of the chair back is discernable, though much distorted, through the main part of the bulb next to the model’s head.

The model’s hair itself is reflected in the adjacent wall of the bulb in just the way a convex surface would present.

Given the multiplicity of lights reflected from the top surface of the bulb, I would guess the setting is either a stage set, or some kind of industrial display in a large building (convention center etc.)

There are two similar, but smaller bulbs, shown at http://scienceservice.si.edu/pages/025061.htm The construction of the filaments and the pin bases are very similar to the bulb in this photo. Those bulbs were described as being for motion picture illumination.

OK, your turn.

Then why do you only see the pattern of the chair at the middle of the bulb? Why don’t you see her arm through the glass? Why do the reflections look so stereotype and show rows of windows? The lady clearly has only one dominant light source.

If anything it’s a flat sheet of glass, that would explain the transparency in some parts of it.

1. The fabric of the chair seat is visibly indented where the electrodes are resting on it.

2. The fabric of the chair back is visible through the bulb envelope.

3. The refracted image of the mantle of the fireplace on the wall behind is visible in the bulb envelope.

4. The position of the model’s fingers are indicitave of holding a 3D object rather than a 2D cutout.

Given the fact that the bulb’s weight is supported by the chair, and the center of gravity of the bulb is almost certainly below the point at which the model is holding it, and the neck of the bulb itself is resting against the arm of the chair, it would not take any effort at all for the model to hold the bulb in that position.

2. and 3. could be faked with airbrushing, but I doubt that it could be done with so convincing an effect.

Most publicity photos from the 1930′s were touched up with an airbrush, the “Photoshop” of the day (to hint at another thread here…)

I have an original photographic print of the Burlington Railroad’s “Mark Twain Zephyr”, which was composited from three different photographs, and under magnification there is not one bit of the Twain that was not touched up with airbrush. It is a commonly-reproduced photo, but without an original print (to get past the typical halftone reproduction) the compositing and retouching would not be evident at all.

There were true artists “in the day”, every bit the equivalent of Photoshop aces today.

It lasts 100 hours?
Less than two weeks at 8 hours a day.
I bet it cost $500.00!