1870.] OSLER — ON CANADIAN DIATOMACE^. 145 



frustules. The true act of reproduction is, comparatively speak- 

 ing, rare among the Diatomaceoc, and probably only occurs when 

 conditions become unf\ivourable to the process of self-division. — 

 In this act two frustules approach one another — their concave 

 surfaces being in apposition ; from each of these surfaces two 

 conical projections of the Endochrome are seen — these coalescing 

 become developed into sporangial frustules, which are considera- 

 bly larger than the parent ones, but exhibit similar markings. 

 Varieties of this mode occur in several genera; in some 

 (Himantidium) the product of the united Endochrome is a 

 single sporangial frustule, while in others (Coccoueis, etc.) the 

 Endochrome of a single frustule escaping, may develop into a 

 sporangium. During the act both old and developing frustules 

 are enclosed in a thick layer of mucous. The subsequent history 

 of the sporangia is as yet very imperfectly understood ; but it 

 is probable that their contents break up into gonidia, and these 

 becoming encysted, develop into several individuals — though 

 some believe that the sporangia undergo self-division, to a limited 

 extent, before breaking up into the gonidia. 



The movements of the living frustule are of a most peculiar 

 kind, and are generally described as a "series of successive jerks 

 in a straight line, and a return, after a slight pause, upon the 

 same path." To explain this motion various hypotheses have 

 been advanced, from time to time. Some observers supposed that 

 cilia were the active agents in producing it, and even went so far 

 as to publish woodcuts of the Diatoms, with the cilia at either 

 end. These have been proved, by Mr. Wenham, to have been 

 "optical delusions." It is true, however, that hair-like processes, 

 uniformly arranged, and bearing a striking resemblance to cilia, 

 are very often seen attached to Diatoms — (this seems especially 

 the case in Nitzshia sigmoida3a, though it not uncommonly hap- 

 pens that the Diatoms of a whole gathering have them) — but 

 these are never seen in motion, and appear to impede rather than 

 assist the movements. They are, in all probability, of fungoid 

 origin. Nageli's hypothesis, namely — that the movements are 

 produced by cndosmotic and exosmotic currents, is one which 

 has met with considerable favour. It was advocated by the late 

 Prof. Smith, and still is, I believe, by Dr. Carpenter. Prof. Max 

 Schultze, of Bonn, has recently advanced a view, which certainly 

 appears reasonable, if borne out by facts. It is this : He 

 supposes that the clear, or slightly granular protoplasmic fluid, 



