They have two purposes. To measure the slates’ lengths and to set out (gauge) the courses so that each slate has sufficient headlap over the slates next but one below - course three overlaps course one by X inches; course four over two etc. The amount of head lap varies with the pitch of the roof, less on steeper pitches; the driving rain exposure of the building, less for less exposed roofs; and the position of the slates up the roof slope, less nearer the top. Sometimes, the headlap is greater on the side of the building exposed to the prevailing wind.

The first thing to understand is that the names of the slates are not their lengths in inches (or any other unit) even if the name is a number. They are traditional sizes and the actual sizes vary from region to region. So a 'sixteen' isn’t sixteen inches long and may be longer in one part of the country than in another. Also the difference between an eighteen and a sixteen isn’t two inches. Rather than indicating specific lengths, they describe the relative position of a particular slate length along the rafter, so the larger number is a larger slate and would be closer to the eaves.

The length of the rule and hence the largest slate which it applies to, depends on the sizes the local delves could make. So a Purbeck rule is longer than most other limestone slates, simply because Purbeck stone-slates come in larger sizes. That said, any region will have extra-large and extra-small slates and these are treated as exceptions or ‘out-rules’. Their actual length varies regionally and reflects the characteristic sizes of the region's slates. 

All the sizes and half sizes are marked except the out-rules. They are usually in two sets - the full sizes and the half sizes - on adjacent faces of the stick. The full sizes are sometimes shown by witness marrks - dots, vees or crosses - to show where you are in the full range of lengths. Different regional names are used for sizes and the are two systems for half sizes. For example in Collyweston and the Cotswolds half sizes are described as long or large versions of the full size. In Purbeck and the Pennines in contrast they are short or scant versions. 

To lay a random slate roof the slates have to be sorted into lengths and the largest placed at the eaves with successively shorter slates laid up to the ridge. There are no rules about how many courses there should be of each length. This is purely the outcome of the mixture of sizes supplied by the delph. (This factor is very important to the local and regional distinctiveness of roofs. Attempts to dictate the gauging for conservation or aesthetic reasons goes against the vernacular style and should always be avoided.)

So the slater’s first job, after the slates have been holed, is to measure their lengths. The effective length of a top fixed slate is the distance from the peg/nail hole to the bottom edge (the tail). The stick is therefore set against each slate in turn with the pin against the tail of the slate, the name (length) is measuread off at the peg hole and the slate placed in the appropriate pile. A set of one length is a parting.

When all the slates are sorted the total width of each length is measured. This figure is then divided by the width of the roof to determine how many courses can be laid in each length. For a gable to gable roof this is simple. For a hipped roof or one with a more complicated plan, adjustments to the calculation have to be made for each course or few courses. 

The slater can now set out the roof for the calculated gauges. Read on - the steps in using a gauging stick.