The theory, for predicting the natural liberation of a mineral from texture measurements on unbroken fragments of ore is complex. Significant contributions in this area have been made by King, Random Fracture Model, Meloy, texture models, and Barbery, integral geometry. Although the theoretical aspects are well advanced, practical application of these methods is difficult, since they require precise measurement on images generated from large fragments of ore. Particularly, a suitable resolution for imaging must be selected that preserves the details of the texture and that completely, encompass the larger features of a single phase in the section. In this work the liberation of monazite from a silexitic ore is predicted using sequences of images acquired from an ore fragment. The images in each sequence were stitched together. The mega-image strips are of high resolution and cover a large enough area that includes the largest features in the texture. Linear measurement is used with advantage, allowing the use of single strips rather than large area patches, The liberation of the monazite is predicted using King's random fracture model, followed by, stereological transformation. The results are compared with liberation measurements carried out on comminuted particles of a rare earth ore of the same procedence.