Despite the seller's description, the figure is of course Diana, not Apollo. It is a fairly accurate copy of authentic lamps depicting this motif that were made in North Africa during the late 2nd to mid 3rd centuries AD. It looks quite convincing at first glance.
However, even judging just by the photos, the unslipped fabric doesn't ring true, the surface looks artificially patinated, and the piercing of the handle is dubiously proportioned. And poor Diana must be slightly alarmed to be reaching up to pluck an arrow from the quiver on her shoulder only to find that the artist has forgotten to include it (though perhaps just as well; she could be a trifle bloodthirsty at times).
Authentic lamps with this motif are quite common. Sadly, fakes of them are even commoner. I suspect this example is a more sophisticated version of F5 and of the same series as F6.
As usual, apart from the dubious assurances of "Thames find" and "British found" by the seller, there is absolutely no indication of where the item came from. A provenance (collecting history) for an item is not only an ethical precaution against inadvertently acquiring looted artefacts; as Elizabeth Marlowe rightly pointed out, it is also an important factor in avoiding fakes. Dealers in fine art have accepted for decades that a work purporting to be by a top artist is worth very little unless accompanied by a watertight provenance. The antiquities market, both high-end and low-end, is every bit as notoriously saturated with fakes as the fine art market. Perhaps it is time that a similar standard for provenance is more widely adopted for antiquities.