Summary. The sealing of obturated root canals which had previously been cleaned chemically by EDTA or mechanically by ultrasound was compared using 60 single‐rooted teeth in an in vitro study. At first, the canals were manually prepared under copious irrigation with NaOCl. Then the specimens were divided into three experimental groups: in group I (control group) the canals were immediately obturated without further preparation; in group II (ultrasound group) the canals were subjected to ultrasound before being obturated, and in group III (EDTA group) the canals were irrigated with EDTA before being obturated. All speciments were then subjected to dye infiltration before being transversely sectioned at various levels from the apex. The amount of leakage was scored on an arbitrary four‐point scale. The results showed some differences in leakage between the three groups at levels close to the apex: EDTA‐treated canals showed the least infiltration, while those treated with ultrasound showed significantly less compared with the control group. The role of the smear layer and its removal is discussed in the light of these results.

Several biopsies of human marginal gingiva were collected from clinically healthy and from slightly inflamed sites. The biopsies from one given site were divided and treated as follows: one portion was immediately fixed and used as a reference; the others were incubated for two hours either in culture medium alone, or in the same medium with different concentrations of elastase purified from human polymorphonuclear leukocytes. The biopsies were then fixed and examined hy electron microscopy. It was found that elastase caused an enlargement of the intercellular spaces of the oral gingival epithelium, interruptions of the basement lamina, and loss of collagen in the underlying connective tissue.Biopsies collected from severely inflamed sites were also studied in the electron microscope. Some of the ultrastructural alterations observed in these inflamed tissues were similar to those experimentally induced by elastase. somal enzymes in vitro in the presence of

Summary. The shaping of curved canals produced by rotational and filing instrumentation techniques was compared in 40 human molar root canals which were prepared in vitro. The tech niques used for filing were the stepback and the serial, and for rotational instrumentation were the ‘balanced force’ and an experimental technique, similar to the balanced force concept, but featuring only clockwise rotated instruments. The results were collected from pre‐ and postoperative radiographs as well as from scanning electron microscope analysis of longitudinally ground half‐sections of the roots. Each preparation was qualitatively rated for its centering, flow, taper, apical design and wall surface appearance. The stepback shapings consistently presented the best taper and apical stop design along with scratch‐free surfaces. The serial technique produced the best centered flow and apically deigned preparations. The rotational techniques produced less consistent results. With the exception of some straightened specimens, the balanced force concept proved to be effective at least for the shaping of moderately curved canals. The lack of taper as well as the absence of apical stop in the balanced force shapings have been found to be more inherent in the concept of shaping than in the technique of the preparation itself. The experimental rotational preparations showed almost exactly the same shaping patterns as the balanced force ones, apart from better taper and apical design but poorer flow and visible instrument scratches. The latter two observations were related to the use of non‐tip‐smoothed instruments.