"Dice mediated augury" Topic

I'm reading a pretty good book (In the Hands of Dante, by Nick Tosches). It has little or nothing to do with gaming, but at one point he describes a method of augury used in medieval or earlier times, in which one rolls dice to pick verses out of the Aeneid, and then tries to interpret them. More or less like picking a verse at random in the Bible, though I never heard of dice being used for that. But no mechanics are given--can anyone think of an elegant way of doing this? I'm running a Roman Empire role play game, so I'd like to add this in.
There are 12 books of the Aeneid, so that's easy, even if I restrict myself to d6s. And typically, editions of the Aeneid have their lines numbered--mine has the English lines numbered, and also has what I assume are the Latin lines quoted in the upper corner of the page. Each book has a different number of lines (Between around 900 and 1265). I imagine I'll have to make up a table and roll against that, but I can't imagine a bunch of ancient wine-bar Italians carrying that sort of table around. Any ideas?

It might be easier in Roman numerals. Some dice might have a blank spot to allow for lower numbers.

First roll 1,2,3 = add 0, 4,5,6 add 6.
Add number of second roll and you have your book. This gives a nice even spread from 1 to 12.

Next roll four times and write the terms down.
Subtract one from each term.
This number is your verse, in base 6. (modulo 6)
You have a number between 0 and 1295, each equally likely.

Say you roll 4362 this becomes = 3251
The right-most column is your ones. = 1
The second column is your sixes. = 5 x 6
The third column is your thirty-sixes = 2 x 6 x 6
Your last column is your 216s = 3 x 6 x 6 x 6
= 648 + 72 + 30 + 1 = 751

Thanks--pretty interesting, if requiring a fair amount of math. One other simple approach--which is a good bit more sloppy--once I've determined the book, I'll divide the pages into 6 groups, roll, pick that set of pages, divide it into 6 groups of lines, and repeat if necessary. Of course, if this really does date back to Roman times, they'd have used scrolls, mostly.
After trying it a few times, "lines" is too small a unit--I need several lines to have anything to work with.

My method looks long, but actually it goes relatively fast. It will be far faster than yours -- once you get used to it.

It's just like using four rolls of a d10 to pick a number between 0 and 9999, except you are using Roman technology.

You can do a simplified version of the I Ching using two eight-sided dice. The math isn't quite the same as the "proper" version, but it's close enough for a quick augury.

In fact, one of the commonly-accepted approaches to the I Ching is to use three coins (any coins, as long as they're the same size) and toss them six times, recording the results.

So I've been looking up dice and knucklebones for purposes of divination. Apparently the Pythagoreans used dodecahedrons for astrology-based divination. 20-siders are known from 2nd Century AD--which puts them a century or so late for the game I'm playing, though the historical record can't be too exact about something like that. Knucklebones look interesting--with a very different probability pattern than dice--but I can't see anywhere I can get ahold of a set.

Very clever, Mr Plant.

Knucklebones look interesting but I can't see anywhere I can get ahold of a set.

link

In New Zealand kids have always played knucklebones, except now they use metal ones -- still shaped like the animal ones though.

I ordered some knucklebones from New Zealand. Thanks for the tip. Though I notice in a statistical analysis of knucklebone results that metal dies produce a different result than bone ones.

And indeed--they arrived and are a bunch of crap, dice-wise. Essentially d2s,instead of biased d4s.