BRITISH DIATOMACEiE. 7* 



of secondary tubes, and their consequent generic separation by ob- 

 servers, who, principally conversant with dried specimens, have re- 

 garded this appearance of importance. 



For the reasons I have just detailed, I am disposed to demur to this 

 course, and to unite all the species under one genus, the name of 

 which must of course be that first proposed by Agardh. I ought, 

 however, to explain why I have not accepted the characters given by 

 Kiitzing as distinctive of his two genera, viz. that the " spermatia " 

 or reproductive bodies are external in Schizonema, and internal and 

 immersed in Micromega. Could these characters be established, a 

 distinction between the two genera must at once be admitted ; but I 

 believe no other observer has been able to substantiate the discoveries 

 of Professor Kiitzing, and my own experience induces me to regard 

 his " spermatia " as parasitic bodies, in no way vitally connected 

 with the fronds on which they occur. In the only species, S. Gre- 

 villii, in which I have noticed conjugation, and the formation of spo- 

 rangia, the process is analogous to that throughout the Diatomaceag ; 

 and the same analogy holds good in the allied genera Colletonema 

 and EncT/onema. In all, conjugation takes place within the frond, 

 or after the dissolution of the mucus-envelope, and follows the usual 

 course. 



The characters hitherto chiefly relied upon in specific descriptions 

 of the Schizonemata also appear to me deficient in permanency, and 

 to lead to a needless multiplication and confusion of species. Neither 

 size, colour, nor ramification is sufficient to distinguish fronds that 

 vary in all these respects with age, locality and exposure. A remark 

 on this subject, made by M. De Brebisson in his * Considerations sur 

 les Diatomees,' 1838, p. 8, is deserving of attention, viz.: — "La plu- 

 part des especes doivent etre revues de nouveau, et etudiees avec 

 plus de soin en ce qui concerne les frustules. On n'a pas tenu assez 

 compte de leurs formes exactes ; les especes ont ete seulement 

 etablies d'apres la disposition et la couleur des frondes ou filamens, 

 caracteres trop variables." Nothing can be more just ; the size, 

 form, striation and structure of the frustules supplying far more cer- 

 tain elements for specific distinctions than the characters hitherto 

 too often relied upon. Acting upon the above suggestion of M. De 

 Brebisson, I divide the species into two sections ; those of the first 

 having frustules firmly siliceous, and fronds in consequence some- 



