Carbonate breccias in beds up to 1-2 m thick interbedded with coarse, reef-derived, carbonate sands and fine, basinal, terrigenous and carbonate muds are present seaward of the southeastern margin of the Devonian Miette Reef Complex of the Canadian Rocky Mountains. The majority of the breccia fragments are cemented carbonate sand set in a matrix of noncemented sand. The allochems of both the fragments and the matrix are very similar in composition. The associated beds of carbonate sand that are

more »

... intercalated with the breccias show similar textures. Within the latter, both boudi-«age-cemented layers and nodules with tension cracks, as well as slumped sequences of cemented sand in a noncemented matrix are seen. The breccia beds are therefore considered to have formed by differential submarine cementation of fore-reef carbonate sands coupled with downslope creep and slumping. Breccias of this type do not record a profound break within the reef complex but correspond to times of maximum supply of sand to the forereef slope. Slumping is the direct result of instability of the sand accumulation on the slope during periods of more rapid sedimentation. HOROWITZ, DANIEL H., Esso Production Research Co., Houston, TX 77001, and HARRY M. BUCH-NER, 12610 Pebblebrook, Houston, TX 77204 DIPMETEE ERRORS AND RELIABILITY Dipmeter data can be used successfully to recognize structural anomalies and may yield valuable stratigraphic information provided one can distinguish valid dips and comprehend their geologic significance. A basic understanding of the techniques used to process the data and of the limitations of the recording tool can assist in achieving this goal. Correlation, the process by which raw dipmeter information is transformed into dips, is the fundamental link between borehole information and geologic interpretation of diplogs, yet many geologists are not fully aware of how it can influence dip results. Because of correlation problems, generalized rather than true formation dips are generated; only if beds are strictly parallel will both be equal. The type of correlation program also determines the extent to which detailed dip changes are masked or revealed. If borehole irregularities are present or the strata are poorly bedded, slumped, or have local heterogeneities, revision of the correlation program may result in data improvement in some cases. Dip errors and inconsistencies are present in almost every diplog. Mechanically induced errors arising from obvious malfunctions or limitations of the dipmeter tool can be detected by proper examination of the field-recorded dipmeter log. Other errors and inconsistencies are less easily detected and generally must be tolerated in the interpretation stage. Comparison of dips from duplicate logging runs is an important tech-nique for estimating the magnitude of these more subtle errors and inconsistencies.