In this study, odontoblasts and predentin of human teeth were stereoscopically observed with the scanning electron microscope. Permanent human teeth were broken with a hammer immediately after extraction and the pulp was removed. The tooth fragments were fixed and dehydrated with acetone and dried at room temperature. The inner pulp cavity surfaces or predentin of the tooth pieces were coated with gold for observation with the scanning electron microscope.
Despite the complete removal of the pulp microscopically, one could see on the pulpal surfaces attached to the predentin clumps of soft tissue representing parts of the superficial layer of the pulp. In regions without adherent soft tissues, there appeared honeycomb-like predentin with innumerable small holes. The remaining soft tissue revealed the accumulation of columnar cells or odontoblasts (Fig. 1). These cells measured about 5 microns in width and 35 microns in height, and were arranged vertical to the predentin surface. Although the surface structure of the odontoblast was not definitely observed, each cell extended as a cytoplasmic process, Tomes' fiber, into a dentinal tubule.
Most of the Tomes' fibers extending into the dentinal tubules from the odontoblasts were flat and found to be adherent to the dentinal tubule walls. However, Tomes' fibers that were not flat were occasionally observed. The surface of the Tomes' fiber was not smooth and its nature could not be adequately defined (Fig. 2). Tomes' fibers are considerably smaller than the diameters of dentinal tubules. This probably represents an artifact developed in the process of preparation of the specimen[3]. Although KUWAJIMA et al. [1] showed smooth Tomes' fibers, the rough appearance of the surface of Tomes' fibers is clearly seen in this observation (Fig. 2). However, the reason for the appearance is yet unknown.
In the predentin the matrix fibers around the dentinal tubules show a basket-like structure. The appearance of these fibers has a marked resemblance to connective tissue fibers of the skin according to FUJITA, TOKUNAGA and INOUE[2]. The direction of the matrix fibers was in agreement with that observed by ICHIJO[3] with a silver impregnation technique.
The matrix fibers ranged from single fine fibrils to bundles of fibrils. They also intertwined with each other (Fig. 2). While matrix fibers were shown to be collagenous by YASUZUMI and OBATA [4], the present study yielded no photographic image of the presence of stripes or cross striations on the surface of the fiber. Further studies are still indicated.
Fibers which are different from Tomes' fibers are frequently found on the surface of the predentin. These fibers, indicated by the arrows in Fig. 1, are long and thicker than matrix fibers. There is an occasional fusion of fibers to each other, some entering into the dentinal tubules and others eventually transforming into matrix fibers. It was not possible to evaluate the correspondence of these fibers to those seen by the light microscope and conventional electron microscope.
(The authors wish to acknowledge their thanks to Dr. Grant VAN HUYSEN, Indiana University School of Dentistry, for his suggestions)

View full abstract

The linear dimensional changes of cylindrical specimens, 12mm high and 6mm in diameter, of twelve brands of silicate cement in distilled water, mineral oil and alternate drying and wetting cycles were observed.
1. Three cements showed expansion in water, and the others showed contraction in water. Cements which expanded in water had comparatively slow setting time.
2. In mineral oil, all cements contracted. Some cements displayed more contraction in oil than in water.
3. The thinner the consistency, the larger the observed change in water. Mixes of standard consistency produced the least change in oil.
4. Cements did not return to original dimensions after alternate drying and wetting.
5. Delayed changes in water for 300 days and in oil for 500 days were observed. The changes in oil took place gradually and continued for a long time compared with the changes in water. The rate of changes in oil will indicate the setting speed or rate of powder/liquid reaction regardless of other factors.

View full abstract