Ross: Classification of Diatoms 401 



It has already been pointed out that stereomicrographs are much more 

 informative than single ones. The techniques for obtaining these and mount- 

 ing them for examination have been described by Little (1958, 1962). It is 

 also important to ensure that the micrographs are taken at a magnification and 

 with a resolution sufficient to show the true structure of the valve. Low 

 power micrographs of the specimen, which will enable its identity to be checked, 

 should also be taken. 



New Observations on Diatom Ultrastrudure 



The species originally described as Schizostauron crucicula Clrun. ex Cleve and 

 S. karstenii Zanon are currently placed in the genus Stanroneis, the structure 

 associated with their central nodule being interpreted as a bifid stauros. Speci- 

 mens of these two species were recently encountered in some gatherings from 

 Lake Tanganyika, and in the same material two undescribed species which 

 seemed related were also found. One of these was very similar to the two 

 known species, but the other had the asymmetry characteristic of the genus 

 Amphora Ehrenb., i.e., both its apical and its pervalvar axes were curved. Al- 

 though it was possible to be reasonably certain under the light microscope that 

 the structures associated with the central nodule were not very similar to an 

 ordinary stauros, details of their form could not be made out with certainty. 

 There was also need to confirm that the asymmetric species differed from the 

 others only in shape and not in any point of structure. Specimens of Schizo- 

 stauron crucicula, S. karstenii, and the asymmetric form were therefore examined 

 in the electron microscope and stereomicrographs of them were obtained. 

 Specimens of the type species of Stauroneis, S. phoenicenteron, and of 5". anceps 

 and S. smithii Grun. were also examined for comparison. These observations, 

 which are reported in detail by Ross (1963), confirmed that the species with a 

 so-called "bifid stauros" were so different from S. phoenicenteron that they 

 should be placed in a separate genus, for which the correct name is Caparto- 

 gramma Kuff. Also, S. phoenicenteron and S. anceps were found to be very 

 similar, but to differ greatly from 5. smithii. The results may be summarized 

 as follows. 



(1) Stauroneis phoenicenteron (figures 1 and 2) and S. anceps (figures 3 and 

 4) have a stauros which is a wide but not very deep thickening of the valve. 

 The chambers that form their striae are elongated along the direction of the 

 stria, especially near the inner surface, where they are separated by a very 

 narrow wall. These chambers are closed on their inner side by a membrane 

 with fine pores in triangular tesselation and on the outer side by a membrane 

 with a broad slit along the direction of the stria. The length of this slit is 

 shorter than the length of the main part of the chamber. 



(2) Stauroneis smithii (figures 5 and 6) has a deep and narrow thickening 

 across the valve. Its chambers are not close; they are approximately circular 

 and are closed on the inner side by a membrane with fine pores in triangular 

 tesselation and on the outer side by a membrane with a narrowly elliptical 

 opening of which the major axis is across the direction of the stria and is longer 

 than the diameter of the main part of the chamber. 



(3) All three species of Capartogramma (for illustrations see Ross, 1963) have 

 on either side of the central nodule two, or occasionally three, deep and very nar- 



