METHODS

Tissues were fixed in buffered formalin and paraffin sections were cut at 10 microns. Sections were stained with hematoxylin and eosin, periodic acid-Schiff, toluidine blue, and orcein-Giemsa. Each specimen was studied completely by means of vertical serial sections.

Tn-Dimensional Models Many of the basal cell growths described in this report were studied by making tracings of serial sections with the Camera Lucida. This device con-sists of a prism situated over the eyepiece of the microscope and an attached mirror which reflects the image of the section downward onto the table next to the microscope. The sections may then be traced onto white paper or a card placed at the point where the image is reflected onto the table. If such tracings are not made, it becomes very difficult to be certain where one proliferation begins and another ends. From these tracings the tumors can be reconstructed as tn-dimensional models. The Camera Lucida drawings are transferred by tracing over carbon paper onto a firm white cardboard-like material called" process board"(obtainable at art supply stores). The process board used was approximately 1 mm. thick, which was the approximate thickness re-quired to make the models proportionate in three dimensions. The process board tracings were cut out with a razor-blade cutting device, glued to-