The answer came from the experiments I’d been conducting, semi-covertly, into why my phone broke. I had no intention of using my laptop or even another phone as a guinea pig, so a quick trip to Computers For Africa, who refurbish abandoned computers and donate them abroad, netted me a bag full of chips and a motherboard that I suspected came from an Atari ST. I used masking tape to set marks at twenty-centimetre intervals along the length of a bench and once I had a chip placed at every mark, I carefully positioned my hand and cast a werelight. The trick in science is to try and change only one variable at a time, but I felt I’d gained enough fine control to produce the same intensity of werelight consistently each time. I spent an entire day conjuring up lights and then checking each chip for damage under the microscope. All to no avail, except for pissing off Nightingale who said that if I had that much time to waste I should be able to tell him the difference between propositions of the accusative and the ablative kind.

Then he distracted me by teaching me my first Adjectivum, which is a forma which changes some aspect of another forma. This Adjectivum was called Iactus, which combined with Impello should, theoretically, have allowed me to float an apple around the room. After two weeks of exploding apples I’d got to the point where I could reliably whoosh an apple down the length of the lab with a fair degree of accuracy. Nightingale said that the next stage was catching things that were thrown towards me, which took us back to exploding apples, and that’s where we were the day the clocks went forward and we paid our respects to Father Thames.

It was while I was in the interview room watching Seawoll gently pluck the facts from Willard Jones’s testimony that I had my breakthrough. Magic, it turned out, was just like science in that sometimes it was a question of spotting the bleeding obvious. Just as Galileo spotted that objects accelerate under gravity at the same rate regardless of their weight, I spotted that the big difference between my mobile phone and the various microchips I’d been experimenting on was that my mobile phone was connected up to its battery when it got fried.