Cocconeis, Cymbella, Amphora and Amphiphora. 5 



cell, and there wrapped themselves up into a small globular 

 form*. 



Again, in the conjugations above described, one frustule is 

 always smaller than the other, as if only a certain quantity of 

 cell-contents were required to make up the portion necessary for 

 the reproduction of the new sporangia; and this is commonly 

 the case also in the conjugation of Sjnrogyra, viz. that one cell 

 is smaller than the other ; while in Claduphora, the whole of the 

 contents of a cell move towards one end of it, where they are 

 condensed into the elliptical form of spore proper to this genus, 

 without any cell-division or conjugation of separate compart- 

 ments. 



Hence it becomes questionable, whether the division of the 

 cell has anything to do with the formation of the spore ; while 

 again it cannot be determined that the formation of the spore is 

 the process ordained for the perpetuation of the species, until it 

 be known in what this process essentially consists, or whether 

 sporing is the only means of reproduction. 



On comparing the si)ore-forniations above mentioned with those 

 described by Mr. Thwaites in the Piatomese, it will be observed, 

 that Cocconeis agrees chietly with Frayilaria pectinalis, whose 

 single sporangium, developed from two small frustules, " is at 

 first cylindrical, soon assumes a flattened, somewhat quadran- 

 gular form, and in many, but not in all cases, undergoes fissi- 

 parous division before it has put on the exact appearance of a 

 frustule of the Fragilaria t-" 



The processes again in Cymhella Pediculus and Amphora ovalis 

 respectively, agree with those of Cocconema lanceolatum % and 

 Eunotia turyida §. In the latter, the way in which the process 

 commences is detailed by Mr. Thwaites, but nothing is said 

 about it in the former ; and although it is easy to conceive how 

 the sporangial frustules in Amphora ovalis should be formed like 

 those of Eunotia turyida, yet it is not so easy to conceive how 

 the processes commence in Cocconema and Cymbella, where the 

 sporangial frustules lie parallel with instead of across the con- 

 jugating ones; unless we assume that another act takes place, 

 in which either the conjugating or the sporangial frustules turn 

 round to obtain this position. 



* Should these small spores develope a filament, the latter would pro- 

 bably be below the uoimal size of the species ; though the original size 

 might be gained by the whole of the contents of a single cell of this deve- 

 lo]>ment, with or without those of the cell of another filament, passing into 

 one new spore. In this way among the Zygnemaceae, as well as among the 

 Diatomea;, species may be added to this family and subclass respecti^ ely 

 which after all are mere varieties. 



t Loc. cit. vol. XX. p. '.i34. 



X Loc. cit. vol. XX. pi. 22 C. fig. 3, &c. § Idem, pi. 1. 



