It is a bit dependant on the exact internals what is the exact cause, but the most likely one is that the system to switch the light on and off uses an electronic element called a Triac. If it can also be dimmed this is very probable even.
It is potentially possible to burn out a Triac by suddenly switching on a load that's a bit over its limit. If the original engineer had foreseen this he may have made a nice circuit that turns on the Triac when the AC voltage curve is at zero, to make sure it doesn't get overloaded. By turning in the lamp you can have accidentally connected it exactly at the highest point in the AC voltage curve, or near it, causing the maximum current to flow very suddenly through a cold lamp. (Cold lamps very shortly require more current to start glowing if it's a filament or even an affordable energy saver lamp).
If this happened the Triac can stay in conduction mode "for ever", requiring repairs by someone who knows what he's doing.
But to be 100% sure some close-and-personal research would be required.
It's possible the panel you have illustrated is one of several such panels serving power into the unit, so you may need to do an exhaustive search for any others.
It complicates things if this building has more than one unit. Wiring to the other units could be still energized obviously... And this could create false readings on test equipment (capacitive coupling) -- or there could be actual cross-connection between the units, so your unit might not be fully de-energized unless others are too.
Cross-connection could be a load in your unit served out of their circuit... Or even a neutral cross, where their load is served from their panel, but it uses the neutral on your panel, and if you sever that neutral wire which has current flowing through it, the upstream side will be energized to 230V.
A lot of testers are complete junk, or are not made for house wiring and need some skill to successfully apply.
A common night light is surprisingly useful. It is simple, honest and not particularly reliable, which means you must check and double-check it. I honestly find them more useful than 3-lamp testers, though those are OK for a final check for new outlets.
Plain old analog voltmeters with the moving needle are useful.
Digital voltmeters (DVM) are useful but a little tricky. They are so sensitive they can pick up micro-currents from capacitive coupling, which happens when a dead wire runs physically parallel to a still-energized wire. So you have a DVM a reading of 109 volts on a 120V circuit that you thought was turned off - plug a night-light in, and the voltage mysteriously disappears.
If the electricity is powerful enough to illuminate any light, even an indicator light -- that's not capacitive coupling, that's genuine power.
Best Answer
The loss of electrical power described in the question may be symptomatic of potentially serious issues that might result in fire and/or injurious or fatal electri shock. It needs prompt attention from a person competent to analyze and work on residential electrical systems. Hiring a licensed electrician is strongly recommended as a general recommendation.
The interruption of power to a portion of your home was undoubtedly caused by an absence of electricity in the conductors serving that portion of the building. Normally this is caused by a open circuit.
It is easy to talk in theory. However, a house that was built in 1890 was certainly built without electrical service. Such service was added later. Because there are both circuit breakers and fuses in the ceiling, it is highly likely that several renovations/additions to the electrical service have been made since it was installed.
Therefore, the only way to determine the pertinent details is by physically observing the actual installation.