.' 



DIATOMACEJS 597 



continued connection of the two frustules by its means gives rise to 

 an appearance of two complete frustules having been developed 

 within the original (fig. 445, A, C) ; subsequently, however, the two 

 new frustules slip out of the hoop, which then becomes completely 

 detached. The same thing happens with many other diatoms, so 

 that the hoops are to be found in large numbers in the settlings of 

 water in which these plants have long been growing. 



But in some other cases all trace of the hoop is lost, so that it 

 may be questioned whether it has ever been properly silicified, and 

 whether it does not become fused (as it were) into the gelatinous 

 envelope. During the healthy life of the diatom l the process of 

 binary division is continually being repeated ; and a very rapid 

 multiplication of frustules thus takes place, all of which must be 

 considered to be repetitions of one and the same individual form. 

 Hence it may happen that myriads of frustules maybe found in one- 

 locality, uniformly distinguished by some peculiarity of form, size, 

 or marking, which may yet have had the same remote origin as 

 another collection of frustules found in some different locality, and 

 alike distinguished by some peculiarity of its own. For then-- i> 

 strong reason to believe that such differences spring up among the 

 progeny of any true generative act, and that when that progeny is 

 dispersed by currents into different localities, each will continue to 

 multiply its own special type so long as the process of binary division 



iioes Oil. 



We have seen that division is of the nature of multiplication, 

 and not of reproduction ; and that, where it does take place, it must 

 be regarded as the exception, and not as the rule. As respects 

 reproduction, Count Castracane, who was an observer during 

 thirty years devoted to the study of diatoms, had the opportunity 

 of noting in what way the process differs in particular cases. 

 He contended that he had been able to see in a Podosphenia the 

 emission of gonids or sporules or embryonal forms, in the same way 

 in which Rabenhorst saw it in Melosira varians, and O'Meara in 

 Pleurosigma Spencerii ; and in another case there were seen a 

 number of oval cysts of a species of Navicula easily recognisable. 

 The greater number of these were in a quiescent state ; but some 

 few were seen in motion by means of two flagelliforin cilia ; so that 

 these larger or smaller cysts represented xygo^pores. and some of 

 them were shown to be zob'zygospores. Castracane had the good 

 fortune to meet with a number of large and small oval cysts 

 imbedded in a gelatinous mass, all of them having in the centre t\vc 

 similar corpuscles. From the condition of two greenish oblong 

 indistinct forms, these went on, by an easy transition, to manifest 

 themselves as naviculoid types, and at length developed into full- 

 grown frustules of Mastogloia. All this proved, in his judgment, 

 how reproduction in diatoms may present itself in different forms 

 and with different peculiarities ; for which reason one ought to 

 avoid arguing from special cases to general laws. The only thing 

 which can be asserted of all cases of reproduction, is that it must 

 be preceded by conjugation, which results in the fertilisation of the 

 1 This refers to those diatoms in which binary subdivision can take place. 



