Ross: Classification of Diatoms 399 



only in Chaetoceros Ehrenb. and Pinnularia of the larger genera is the proportion 

 studied greater than in Navicula. 



There are two parts of the system of classification of the diatoms in which the 

 currently accepted taxonomy above the specific level is patently unsatisfactory: 

 the famiUes Navicuiaceae and Biddulphiaceae. In both, species are grouped 

 with others to which they seem only distantly related and separated from those 

 which seem close to them. In a taxonomic investigation of a small group 

 of species in the Navicuiaceae on which I was recently engaged, I decided 

 that electron stereomicrographs would be useful in elucidating a particular 

 point about the structures connected with the central nodule. Through the 

 kindness of K. Little of the Nuffield Orthopaedic Centre, Oxford, England, who 

 is responsible for all the micrographs illustrating this paper, these were obtained. 

 They showed not only the details of the central structure but also the ultra- 

 structure of the perforations through the valve, and the correlations between 

 these two suggested that ultrastructure might well form a guide to a revision 

 of the limits of Stauroneis Ehrenb. and possibly certain other genera, and that 

 the attitude of Hustedt (1959), Hendey (1959), and Lund (1962) to its use for 

 this purpose was unduly defeatist. These observations are being extended, and 

 much more needs to be done before any firm conclusions can be reached. This 

 paper cannot, in consequence, be anything more than a report on progress to 

 date, but its object will be fulfilled if it dissipates doubts about the value of 

 ultrastructure as a source of taxonomic characters and stimulates others to 

 work on similar lines. 



Technique 



This approach necessitates the accumulation of electron micrographs of a 

 large proportion of the species in the group under investigation. Many species 

 of diatom occur only as comparatively rare members of the assemblage con- 

 tained in a particular gathering. To obtain the electron micrographs needed 

 in a taxonomic investigation accordingly demands the use of a techniciue similar 

 to that used in the making of selected slides of individual specimens for the 

 light microscope. Reliance on serendipity, which has hitherto been the normal 

 practice when choosing specimens for investigation with the electron micro- 

 scope, will not suffice. 



Each worker who makes selected slides of individual diatoms develops a 

 technique which suits the resources of his own laboratory and his personal 

 characteristics, in particular the steadiness of his hand. This account of the 

 method I have used for selecting individual diatoms for study with the electron 

 microscope, which is based upon that which I use when making selected slides 

 for examination with the light microscope, should be taken only as a general 

 guide and not as a model to be rigidly followed in all of its details. 



One starts with a suspension in distilled water of chemically cleaned diatom 

 frustules (for methods see Hustedt, 1927, 1958, Swatman, 1937, Hendey, 1938, 

 1951, Leboime, 1952, van der Werff, 1955, Barber, 1962) which is known to 

 contain the diatom which it is desired to study. A few drops of this are allowed 

 to evaporate, preferably on a mica surface, to which diatoms adhere less than 

 they do to glass. Heat should not be used as convection currents cause the 

 diatoms to clump together. Diatom frustules apparently adsorb some of the 



