For a number of years I was proud to count the late BBC Radio Wiltshire presenter Mark O'Donnell as a friend. He very sadly died at the end of 2019 - but I wanted to look back and remember some remarkable inserts we made for his show.  The idea was to visit science and technology locations in Wiltshire. It was a great opportunity to find out more about these places, whether historical or modern. What made the visits was Mark's warm presence and ability to draw the listener in to the scene. Sadly the BBC has not kept the tapes of the visits, but here are a few written highlights from me. The startling significance of Mr Talbot's spectacles - visiting Lacock Abbey and the scene of the first negative image A revelation in Wroughton - at the Science Museum controlled environment store and library (a location that so impressed me that I suggested it as a location for my little TV piece teaching quantum physics to then BBC business editor Robert Peston ) How oxygen was first di

It’s always satisfying to come across a well-written murder mystery and discover there are several more waiting. In this case I accidentally started with book 5, but have since gone back to the first of Tim Sullivan’s novels featuring DS Cross. The most interesting feature of the series is that Cross is on the autism spectrum. This gives him some distinctive advantages over his colleagues, while also offering some challenges. On the whole Sullivan handles this well - in this book, almost all of Cross’s colleagues regard him with affection, though we are told that in the past he was treated badly. That is perhaps the most unlikely aspect - it’s hard to imagine that policing has so many suitably thoughtful officers, though I may be resorting too much to stereotype. The murder victim is a Catholic monk, with much of the action taking place in an abbey - also well handled and providing a neat tie-in to Cross’s enthusiasm for church organs.  The monk’s background is unusual, giving opportun

The other day, reading Tom Chivers' excellent book on Bayesian statistics  Everything is Predictable , I was reminded of that old chestnut, 'absence of evidence is not evidence of absence.' This is often put forward as if it were a powerful logical argument. But, in reality, it's a bit of common sense that sometimes works, but always oversimplifies. In case you aren't familiar with the expression, I might say that I've never seen any evidence that dark matter exists (as opposed to the behaviour of galaxies and galactic clusters attributed to dark matter), but I shouldn't take that as evidence that dark matter doesn't exist. As Tom Chivers points out, this is very frequentist thinking. The Bayesian approach would be that every good quality experiment that fails to find dark matter modifies our priors - it can be used to reduce the probability that it exists. Interestingly, this somewhat trite saying only tends to be wheeled out when responding to a theor