| Abstract: | The scanning electron microscope has provided much information about the structure of the valves of centric diatoms that supplements or, at times, corrects the accounts based on transmission electron microscopy. In particular, the morphology of the special features that have been called processes, spines, ocelli, tubules, etc., has been elucidated. This paper reviews the results so far obtained, discusses the way in which the various structures found should be grouped, and proposes a terminology for the description of these. As far as has been possible, existing terms have been retained and used in their current sense, but precision has not always permitted this; definitions of all the terms adopted are given. The concept that diatom frustules could be divided into those with a one-layered structure and those with a two-layered structure is rejected. |
