The characteristics of directional discontinuities (DDs) in the interplanetary magnetic field have been studied using data from the Mariner 10 primary mission. The entire data set was surveyed using an automated procedure to identify DDs as changes in field direction of at least 30¿ in a 42-s interval. This study yielded an r-1.3¿0.4 heliographic distance dependence for the daily average number of discontinuities per hour. In addition to this statistical survey, DDs were visually identified using 1.2-s averages for three selected time intervals, and the corresponding 40-ms data were studied in detail by means of the Sonnerup-Cahill variance procedure.

The three time intervals were November 4--17, 1973, February 5--12 and April 3--10, 1974, corresponding to the equally spaced heliocentric distances of 1.0, 0.72, and 0.46 AU. Editing the candidate DDs produced a total of 644 events. Two methods were used to estimate the ratio of the number of tangential dicontinuities (TDs) to the number of rotational discontinuities (RDs). In the first approach those DDs with substantial normal components (Bn/≥0.3) were interpreted as RDs, and the remainder were considered to be TDs, except that some RDs were eliminated on the basis of unacceptably large relative magnitude variances across their discontinuity zone.

The second method considered the total number of RDs to be the sum of those DDs with substantial normal components (again excluding those with large variances) plus an estimated number of those DDs with small normal components. The estimate was based of the assumption that there is a uniform distribution of RDs per degree of discontinuity cone angle &bgr; (=cos-1 ‖Bn‖/B, where B is the average magnitude of the field across the DD) for all &bgr;. Then all other DDs are assumed to be TDs.

Both methods showed that the ratio of TDs to RDs decreased with decreasing radial distance, being 1.5, 1.2, and 0.86 for the distances 1.0, 0.72, and 0.46 AU, respectively, from the first method and 0.89, 0.66, and 0.43, respectively, from the second (preferred) method. A decrease in average discontinuity thickness of 41% was found between 1.0 and 0.72 AU and 56% between 1.0 and 0.46 AU, independent of type (TD or RD). This decrease in thickness for decreasing r is in qualitative agreement with Pioneer 10 observations between 1 and 5 AU. When the individual DD thickness at the three locations given above was nearly constant, 36¿5 RL. This held true for both RDs and TDs separately.

Statistical distributions of these and other properties, such as normal directions and discontinuity plane angles (&ohgr;), are presented. No obvious relationship was found between &ohgr; and the thickness of either TDs or RDs when widely separated locations were examined.