O'Meara — Report on the Irish Diatomacew. 243 



The reproductive process has not been observed in more than about 

 sixty-five species, and in some of these cases there is a difference of 

 opinion as to the facts. A system, therefore, in which this process 

 ■constitutes an important part of the ground plan, is practically ob- 

 jectionable, as founded on hypothesis. 



The system which appears to me to have most to recommend it is 

 that which has been matured by Heiberg, founded on the symmetri- 

 cal or unsymmetrical structure of the frustules in their various aspects. 

 There are two principal aspects in which a diatomaceous frustule may 

 be regarded — the front view, in which the hoop or connecting band is 

 presented to the eye, and the side view, in which one or other of the 

 two valves is under observation ; and in both these positions the 

 longitudinal and transverse axes are to be considered. If in these two 

 positions, and in these varied views, exact symmetry obtains, the frus- 

 tule is said to be symmetrical in all its aspects ; but if the two oppo- 

 site valves are not uniform, or the portions of the valves on either 

 side of the transverse or longitudinal axis, on side view or front view, do 

 not exhibit the same proportions or outline, the frustule is said to be 

 unsjnnmetrical on that view or axis on which the difference of form is 

 observable. Such is the ground plan of Heiberg's systematic arrange- 

 ment, and which I have adopted in the present report. It is not, 

 indeed, wholly free from the objections to which other systems are 

 liable, and, possibly, may be open to others peculiarly its own ; but 

 still the principle on which the arrangement is based commends itself 

 as being at once most simple, most comprehensive, and most easily 

 applied. 



There are, however, two very important yet subordinate features 

 of Heiberg's system, in regard to which I cannot adopt the views of 

 that distinguished author. He ranges the numerous Cuneate species 

 as aberrant varieties of the families which in other respects they most 

 closely resemble ; for example, Meridion and Asterionella are asso- 

 ciated with the Pragilariese, under the distinctive appellation of 

 Fragilariese cuneatse ; Podosphenia with the Striatillete, as Striatillese 

 C'uneatre ; Gomphonema and Cocconeis with the I-Taviculese, as IS^avi- 

 ouleas cuneatfe ; whereas I have collected the numerous species with a 

 cuneate outline, under the one general group of Cuneatte. Again, 

 the numerous species of Liatomacese, as is well known, exhibit various 

 normal phases of growth. Some species are normally free, others 

 attached by a short gelatinous cushion, or a larger or shorter stipes ; 

 the frustules in some genera are simple, while in others, after self- 

 division, they remain in concatenate or ribbon-like filaments ; in 

 some genera the frustules are naked, while in others they are en- 

 veloped in mucous fronds, of which some are indefinite, others definite, 

 fonning simple or composite tubes. These peculiarities of growth 

 Heiberg treats as of little significance ; and accordingly, the genera 

 Schizonema, Berkleya, and CoUetonema, the frustules of which are 

 included in tubes ; Dickiea, the frustules of which are imbedded in 

 ii less definite mucous mass : Diadcsmis, the frustules of Avhich are 



