Midnight Riot
I came across this book when
james_nicoll and
sleary both posted about the cover!fail. Short version: the main character is a mixed-race cop in modern London, and it's a point in the characterization that he's obviously non-white enough that people tend to edge away from him a bit in the Tube and such. The covers depict him in silhouette ... and the
original, pre-release, covers didn't. For more detail, go to
the original SF Reviews.net forum post.
Anyway.
sleary quoted from an interview with the author that made me interested ... and as I'm almost completely burned out on urban fantasy, that takes quite a bit of doing. Aaronovitch said that he wanted to write a police procedural in the British mode. It was enough to make me download the sample, and after reading it, I promptly bought the book.
Loved it. I've only restrained myself from buying the sequel and starting it (until today) because I know from hard experience that if I read too much of a series in one go, I'll burn out on it, and I want to be fresh for when book 3 comes out, later this year. :)
There are no sexy vampires or werewolves in this book. A giant point in its favor. Peter Grant is a London beat cop marking his time until he gets promoted and reassigned, when he accidentally falls into a murder situation that involves the
other side of London, the mythic and supernatural side, and ends up working with the division of the Metropolitan Police that takes care of it.
The British title is
The Rivers of London, and the rivers are indeed characters, or at least the mythic archetypes are. Most of whom have updated themselves over the years, or mortals have become the archetype of the river.
Aaronovitch also pulls another trick off: Grant uses scientific methods of exploring magic and it does not suck all the whimsy out of it for me. Admittedly, I didn't go into the book for the numinous quality of the magic, but for the police procedural, but I wasn't irritated by Grant's attempts to winkle out the laws behind magic the way I usually am. (Probably because Aaronovitch made it messy, like real science is!)
Quite good, and recommended.