398 Annals New York Academy of Sciences 



rates pores, open at both ends, from chambers, with a septum at one or both 

 ends, and classifies these according to the position and type of perforation of 

 the septum or septa. 



The Use of Diatom Ultrastructure in Taxonomy 



In spite of the large amount of information available about the ultrastructure 

 of diatoms, it has until now been of little use in their taxonomy. Hustedt 

 (1952, 1955), in the course of an interchange of opinion on the subject with 

 Kolbe (1954, 1956), maintained that ultrastructure is more uniform than the 

 features that can be seen with the light microscope, and that its variations 

 show no correlation with the characters used to distinguish genera; ultrastruc- 

 ture, accordingly, cannot be regarded as having any taxonomic significance 

 above the specific level (Hustedt, 1959, pp. V-VI). Hendey (1959) came to a 

 similar conclusion, but added that when a large number of species have been 

 examined it may be possible to subdivide the genus NavicuJa Bory. Views 

 similar to Hustedt 's are presented by Lund (1962) in his recent review of the 

 criteria adopted in classifying algae. 



It is probably not an unfair generalization to suggest that taxonomists are 

 conservative in their outlook, especially in their views about which characters 

 are important in classifying a particular group. They do not seize every 

 opportunity of using a newly discovered set of characters to produce a new 

 system supplanting the current one. They tend rather to keep alterations to 

 a minimum, apart from the addition of numerous new species and taxa of lower 

 rank. One of the most gratifying results of the study of diatom frustules with 

 the electron microscope has been that it has brought to light nothing really 

 surprising. Structure too fine to be resolved with the light microscope has been 

 demonstrated, but this was only to be expected. Nothing which could be seen 

 with the light microscope has been found to have a structure markedly different 

 from that which it was thought to have. This represents a great tribute to the 

 skill and acumen of those who used the light microscope at the limit of its 

 potentialities to elucidate the structure of the diatom valve, especially O. Miiller 

 (1889, 1895, 1896a, b, 1898, 1899, 1900, 19()la, b, 1909) and Hustedt (1926o, 

 h, 1928o, b, 1929a, b, 1935a, b). On the other hand it has meant that no re- 

 visions of the system have been forced upon diatom taxonomists, and in the 

 absence of any such pressure they have not actively pursued the question of 

 how far knowledge gained with the electron microscope could influence classifi- 

 cation above the specific level. 



Although conservatism has played its part in persuading diatomists that 

 ultrastructure can only play a minor role in the taxonomy of the group, they 

 have been helped to reach that conclusion by two other factors. Both of 

 these have already been discussed; they are the inadequate number of species 

 investigated with the electron microscope and the inadequate information 

 about the ultrastructure of many of those examined. Thus, Hendey's (1959) 

 list of the diatoms investigated with the electron microscope includes only 28 

 identified species of Navicula, out of at least 1000 at present known, and it is 

 probable that the information about the structure of many of these is as inade- 

 quate as that which he gives about Stauroneis anceps and S. phoenicenteron 

 (cf., p. 397). For all other genera fewer species have been investigated, and 



