Phylum Phaeophyta [ 73 



the process, but the nuclear cavity remains distinct until the chromosomes have 

 reached the ends of the spindle. The nuclear sap and the spindle are then absorbed 

 by the cytoplasm, but not until the spindle has budded off a new centrosome from 

 each end. 



Subsequent authors, as Karsten (1900), Geitler (1927), Iyengar and Subrahman- 

 yan (1942, 1944), and Subrahmanyan (1947), have not seen as full a series of stages 

 as Lauterbom did. They have found centrosomes in at least some diatoms, and have 

 confirmed the point that the spindle is a cylinder which is surrounded by the 

 chromosomes instead of including them. 



The same authors have described sexual processes in Surirella, Cymbella, Coc- 

 coneis, Cyclotella, and Navicula. In Surirella saxonica as described by Karsten, pairs 

 of the wedge-shaped cells become attached by little bodies of slime at the narrow 

 ends. Each nucleus divides twice, producing four, of which three are digested by the 

 cytoplasm. The two protoplasts then move in amoeboid fashion out of their shells 

 and they and their nuclei unite. The zygote protoplast grows to a size much greater 

 than that of the parent cells and secretes a membrane which becomes silicified. The 

 resulting cell is called an auxospore. 



In most kinds of diatoms, each cell produces two gametes. In some, the cells pair 

 and proceed to produce auxospores individually, without conjugation. Karsten sup- 

 sposed the latter examples to represent a stage in the evolution of sexual reproduc- 

 tion under some zwingender Nothwendigkeit: much more probably, they are pro- 

 ducts of degeneration. In Cyclotella, Iyengar and Subrahmanyan found the produc- 

 tion of auxospores to involve autogamous karyogamy: the nucleus of a solitary cell 

 undergoes meiosis; two of the haploid nuclei are digested, and the two which remain 

 fuse with each other. It is evident that all diatoms are diploid in the vegetative 

 condition. 



The filamentous green Heterokonta Tribonema and Bumilleria are closely similar 

 to the diatom Melosira, and it may reasonably be supposed that they represent the 

 evolutionary origin of the group. 



Diatoms are preserved for study by violent methods which destroy the protoplasts, 

 and the classification is based strictly on characters of the shells. So uniform is the 

 group that Schiitt (in Engler and Prantl, 1896) treated it as a single family. He pro- 

 vided an elaborate subsidiary classification involving two main groups. Subsequent 

 scholars have found his system essentially sound as a representation of nature, but 

 have raised the main groups to the rank of orders and the minor ones in correspond- 

 ing degree. 



Order 1. Disciformia [Disciformes] Kiitzing Phyc. Germ. 112 (1845). 

 Order Appendiculatae Kiitzing 1. c. 

 Centricae Schiitt in Engler and Prantl Nat. Pflanzenfam. I Teil, Abt. lb: 57 



(1896). 

 Order Centricae Campbell Univ. Textb. Bot. 90 (1902). 

 Order Eupodiscales Bessey in Univ. Nebraska Studies 7: 284 (1907). 

 Diatoms basically of radial symmetry, which, however, is often distorted; not 

 motile in the vegetative condition; plastids numerous in the cells. 



These are the more primitive diatoms. The majority are marine. Three types of 

 reproductive cells are known to be produced by them. 



Occasionally, in mass catches of material from the ocean, diatoms are found 

 whose protoplasts have undergone repeated division within the shell and produced 



