DIATOMACKJ-: 



6oi 



subdivision (No. 3, a, b, c), the cells of the new series thus developed 

 presenting the character of those of the original filament (1), but 

 greatly exceeding them in size. From what has been already stated, 

 it seems probable that a gradual reversion to the smaller form takes 

 place in subsequent subdivisions, a further reduction being checked 

 by a new formation of zygospores. The various modes of formation 

 of auxospores in the Diatomacese are classified by Klebahn under 

 five different heads, viz. : (1) Rejuvenescence of a single cell, accom- 

 panied by an increase in size ; this is the simplest type, and one 

 of the most common. (2) Two daughter-cells are produced from the 

 protoplasm of a mother-cell, and from these arise two auxospores 

 (Achnanthes longipes, Rhabdommcf arcuatum). (3) Two cells lying 

 side by side cast off their old valves, and each grows into an 

 auxospore, without any previous fusion, or any visible interchange 

 of contents ; this is the commonest type of all. (4) A true conjuga- 

 tion takes place ; the protoplasmic contents of the two cells fuse 



FIG. 447. Self -conjugation (?) of Melosira italica (Aulacosira crenulata 

 Thwaites) : 1, simple filament ; 2, filament developing auxospores ; a, &, c, succes- 

 sive stages in the formation of auxospores ; auxospore-frustules in successive stages, 

 a, b, c, of multiplication. 



together into one, and this mass grows into an auxospore. (5) Before 

 conjugation, the protoplasm of each of the two cells divides before- 

 hand into two daughter-cells, and two auxospores are formed by 

 the fusion of a daughter-cell from each mother-cell with the 

 daughter-cell of the other one lying opposite to it ; this is the most 

 complicated process (Amphora ovalis, Epithemia Argus, Rhopa- 

 lodia gibba, &c.). 



The most curious phenomenon presented by diatoms is un- 

 doubtedly their power of movement, which induced Ehrenberg and 

 the other early observers of these organisms to place them erro- 

 neously in the animal kingdom, although it affords no evidence of 

 consciousness. This power of movement, if not common to all 

 diatoms, is very evident in those species which are normally or 

 accidentally free, and most conspicuously in oblong forms, such as the 

 species of Navicula. In those also which are stalked it has been 

 noticed that if, from any cause, a frustule becomes detached, it is 



