492 ANIMAL NATURE OF DIATOMEjE. 



or class, it assists us greatly if we select those that 

 present the greatest complication of structure, because 

 the comparison of these may reveal the organisation 

 which others disguise under an assumed simplicity. 



In the three tribes we have examined (Striatse, Vittatse, 

 Areolatse), Kiitzing arranges all the Diatomese which he 

 believes to form a section of the first of the two classes 

 into which he divides all the Algse. No one who con- 

 siders that even the Charge are comprised in the same 

 class, will feel any wonder at this assemblage. But 

 without calculating either the degree or the position 

 attributed to the group of Diatomese in the Organic 

 Kingdom, and attending only to the examination of their 

 ulterior division, I believe that the observations adduced 

 may serve to demonstrate that in the three proposed 

 tribes we have unnatural dismemberments and asso- 

 ciations. The same conclusion prevails also in respect 

 to the six orders (Striatse, astomaticse and stomaticse; 

 Vittatge, astomaticse and stomaticse ; Areolatse, disciformes 

 and appendiculatse,) as well as to the ulterior divisions in 

 the first two, taken from the continuity or interruption 

 of the striae and the presence of one or two stomatic 

 apertures. Much more naturally established do we find 

 the nineteen families, (Eunotiese, Meridieae, Fragillariese, 

 Melosirese, Surirellese, Cocconoidese, Achnanthese, Cym- 

 belleae, Gomphonemeae, Naviculese, Licmophoreae, Stria- 

 telleae, Tabellarieae, Cosconidiscese, Anguliferae, Tripo- 

 disceae, Biddulphieae, Angulatae, Actinisceae ;) respecting 

 these we have seen that, in the actual state of science, few 

 changes would be fully justified. Such changes appear 

 to me to be the following : — the union of Meridieae with 

 Pragillarieae ; the separation of the genus lAcmophora 

 from the family that bears its name, to join it with 

 Bacillarios and Synedrcs in the family of Surirelleae ; 

 to examine it again comparatively with that of Navi- 

 culeae, to decide upon the presence, constancy, and value 

 of the character of a median aperture, which alone serves 

 to separate them ; the comparison, under this point of 



