The spatial relations of joints and, in particular, structural details of individual joints, offer clues of their origin. Important features of joints have been largely neglected in previous joint studies and the present study is an attempt to determine more closely the true nature of joints in sedimentary rocks and to suggest a mode of origin more in line with field observations than is present theory.

The study area comprises about 2,000 square miles of the Colorado Plateau in northeastern Arizona and southeastern Utah where sedimentary rocks ranging from Pennsylvanian to late Cretaceous in age are exposed.

A simple, non-genetic joint classification is presented based on the spatial relations of joints and the plumose structures on joint faces. Joints are grouped as systematic or non-systematic with cross-joints defined as an important variety of non-systematic joints.

Plumose structures on joint faces indicate that joints are initiated at some structural inhomogeneity within the rock and propagated outward, thus precluding movement in the direction of the joint faces at the time of formation. Spatial relations of systematic joints point to formation at or near the earth's surface in a remarkably homogeneous stress field.

The regional joint pattern is composed of a complex series of overlapping joint trends. The pattern as a whole extends through the entire exposed rock sequence. Intersecting joint trends have no visible effect on each other and may terminate independently in any direction. Each joint trend of the regional pattern crosses several folds of considerable magnitude but does not swing to keep a set angular relation to a fold axis as the axis changes direction.

Hypotheses stating that joints are related genetically to folding are rejected for the mapped area. The shear, tension, or torsion theories of jointing require that only one or two sets of joints be considered as the result of a particular stress condition. Where more sets are present, different stress conditions must be postulated for each set or pair of sets believed to be related genetically. The joint pattern in the mapped area can not be interpreted in such terms without making these assumptions, Alternatively, in accord with theoretical and experimental evidence, semi-diurnal earth tides are considered as a force capable of producing joints in rocks through a fatigue mechanism. Field observations suggest that joints form early in the history of a sediment and are produced succes ively in each new layer of rock as soon as it is capable of fracture. The joint pattern in pre-existing rocks may be reflected upward into new, non-jointed rock, and so control the joint directions.

Much critical data from areas with different geologic histories are needed before a quantitative evaluation of this hypothesis can be made. The question of the ultimate origin of the regional joint pattern and its genetic relation, if any, to other structure at depth can not be answered on data now available.