The Cloudspotter’s Guide is exactly what it says on the tin: a systematic guide to the main types of clouds, their subspecies and varieties, the physics leading to their creation and disappearance, etc. It’s all that, plus a love for clouds. The science is mixed up with all sorts of anecdotes and asides, personal reflections and observations about clouds in art and culture.

This is a pleasant diversion, a charming book. The style is very personal and chatty. This makes it an easy read but at the same time makes it hard to remember many details. There were so many disparate facts that I have already forgotten most of them, and the clouds all blur together again. But nevertheless, it was a pleasant read.

My one complaint is the lack of pictures. Each chapter starts with a nice woodcut-style illustration of a particular type of cloud, but apart from those ten images, and a few colour plates in the middle, there are mostly small black and white photos, flat and grainy – probably because of cost issues. This should be a glossy book with pictures all over the place. Not a coffee-table book, mind you: I rather liked reading it during my commute, and being able to look up from the book and gaze at the clouds above me.

Amazon UK, Amazon US.

Every other fantasy book and RPG, and pretty much every fantasy book or game where a significant part of the action takes place in a major city, has a Thieves’ Guild, possibly a Beggars’ Guild, and an Assassins’ Guild.

Now I can understand the reason and rationale for the first two. Thieves and beggars could well want to prohibit outsiders from crowding onto their turf, divvy up the city, and impose other rules. But I really have trouble understanding how an Assassins’ Guild could possibly exist.

Quite apart from such mundane concerns as “wouldn’t it be rather dangerous for someone to be a registered assassin?”, there is the issue of volume and sustainability. Let’s think about the numbers.

How many members does a guild need to have in order to be called a guild and not just a bunch of guys? A few dozen at least, I’d think. Let’s say 30.

How often would an assassin need to kill someone in order to keep his skills sharp? Once a week seems like a reasonable minimum. That’s 52 kills per assassin and year, and 1560 kills for the entire guild.

How many people in a city? Most fantasy takes place in a late medieval or renaissance-equivalent era. The top 10 cities in this world in the year 1500 ranged from 150,000 inhabitants to 670,000 inhabitants (About.com). Let’s assume, generously, that the cities in the books are really grand ones, say 300,000 souls.

Assuming a life expectancy of 30 years (Wikipedia) and a stable population, there would be 10,000 births and 10,000 deaths every year. Of those 10,000 deaths, about 3,000 would be children in their first 5 years of life. That leaves 7,000 deaths of other causes.

1560 assassinations out of 7,000 deaths would mean that one death out of every 4 or 5 is an assassination. Or to put it another way, deaths from all other causes would need to be 20% fewer than in our average medieval city, or else the city will be emptied pretty quickly.

If we reduce the assassins’ activity level to one kill per month, that’s 360 kills per year or rougly 5% out of all non-infant deaths, which is rather more reasonable. But it means that, on the one hand, the guys wouldn’t get much practice, which means they would not be as skilled, so they would get lower pay, and they would need to have day jobs. And once you’re spending 90% of your time working as a messenger, thief, horse trainer, or whatever, and only killing someone once a month, can you really call yourself an assassin still?

PS: For more medieval demographic calculations, try Medieval Demographics Made Easy.