ULTRASTRUCTURE RESEARCH AS AN AID IN 

 THE CLASSIFICATION OF DIATOMS 



R. Ross 



British Museum {Natural History), London S.W . 7, England 



Present Knowledge of the Ultrastrudure of Diatoms 



The frustules of diatoms were among the first biological objects to be exam- 

 ined with the electron microscope (Mahl, 193^), and in the preceding 20 years 

 a large number of works dealing with the subject have appeared. These have 

 been listed comparatively recently by Hendey (1959), and at the time at which 

 he wrote information about some 300 species was available. Although it is not 

 important to review the results of these studies in detail, there are two points 

 about them which need to be emphasized here. The tirst is that none of this 

 work has been done with any particular taxonomic problem in mind. For the 

 most part it would seem that investigators took the material which came readily 

 to hand, mounted drops of it on electron microscope grids, and took pictures of 

 the forms they found there. This has on occasion led to doubt as to the true 

 identity of the species studied, as in the case of the illustrations published by 

 Kolbe (1951, plate 2, tig. 4, plate 3, figs. 5 and 6) as Navicula subtUissima Cleve, 

 but said by Hustedt (1952, 1955) to be of Anomoeoneis exilis (Kiitz.) Cleve or 

 A. serians var. brachysira (Breb.) Cleve (Kolbe, 1954, 1956, 1959). A more im- 

 portant consequence, however, is that there is not any group of supposedly re- 

 lated species of which more than a small proportion have been studied with the 

 electron microscope. In no case do we know the patterns of similarity and 

 difference and the range of ultraslructure to be found within a single genus, 

 with the possible exception of Pinnularia Ehrenb., of which electron micro- 

 graphs of some 15 species in a genus totaling at least 250 suggest that the 

 ultrastructure is as uniform as that revealed by the light microscope. 



The other important point is that the interpretation of electron micrographs 

 is by no means easy, and also that, in some cases, those published do not give 

 an adequate picture of the structure of the species illustrated, either because the 

 specimen was damaged in preparation or because the resolution is insutlicient. 

 Interpretation is ditlftcult because of the great depth of focus of the electron 

 microscope and the considerable opacity of silica to electrons. Even in pic- 

 tures of complete frustules, the whole is equally in focus. In the light micro- 

 scope it is possible to build up a picture in depth from a series of optical sections 

 obtained by alterations of focus, but this technique is not available to the elec- 

 tron microscopist. When more than one layer is visible it is often not possible 

 to tell from single pictures which lies above which. Much of the valve, also, is 

 completely opaque to electrons, and where this is so there is no information 

 about differences in thickness from differences in transmission of electrons. 

 Stereomicrographs accordingly provide much more information than single 

 prints, as may be seen from the large number published by Helmcke and Krieger 

 (1953, 1954, Helmcke et al., 1961). These authors have applied stereogram- 

 metric techniques to the study of their stereoscopic pairs and have produced 

 models of the structure of a number of species, thus obtaining the maximal 

 amount of information from the data recorded on the micrographs. 



396 



