BRITISH DIATOMACEvE. 11 



although the frustules which compose them may present great diversity 

 in the outline of their valves. Nor in the latter particular do we find 

 any very satisfactory or constant peculiarities ; filaments whose frus- 

 tules are composed of valves of a linear form, as in H. Soleirolii, being 

 frequently found intermixed with others whose valves have terminal 

 constrictions, as in H. pectinale, or one or more dorsal ridges, as in 

 H. undulatum. Under such circumstances, it is difficult to say whether 

 such forms are distinct species, or only varieties whose characters 

 depend upon local or accidental circumstances. If regarded as of 

 specific value, it is evident that the living condition of the plant will 

 afford no certain means of identification, as the side view can alone 

 be ascertained by an examination of frustules when the connexion of 

 the valves has been dissolved ; if, on the other hand, these diversities 

 be esteemed as accidental variations, it will be equally difficult, in the 

 case of fossil or prepared specimens, to determine from the valves 

 alone whether we are dealing with one or several species. A wide 

 examination and comparison of species in both conditions induces 

 me to believe that the number has been unduly increased; and 

 although, in deference to the authority of former observers, I have 

 hesitated to curtail the list further than I have given it below, I am 

 far from certain that (excluding H. Williamsonii as anomalous in 

 form and habitat) the entire number of our native species might not 

 be reduced to two, of which I should regard H. pectinale and H. 

 Arcus as the typical forms ; — the first characterized by a tenacious 

 filament, and the second by the fragility of the connexion uniting 

 the frustules, which are rarely found combined into a lengthened 

 band. To the first type might be referred H. undulatum, Soleirolii, 

 and H. minus, Kiitz. To the second, H, bidens, gracile, and majus, 

 H. exiguum, Breb. (=Eunotia gracilis, W. Sm.), H. denticulatum, 

 Breb., the forms of H. Arcus described by Dr. Gregory under the 

 names of Eunotia bigibba and Himantidium bidens, with numerous 

 other varieties of outline which might be added to his list, together 

 with many species of Eunotia which hold an intermediate place 

 between the two genera, and are occasionally found in fragmentary 

 filaments of from 2 to 8 frustules. The following list of species 

 must therefore be considered as extended somewhat beyond the 

 judgment of the author, to include the varieties described by other 

 writers, while the figures given must be taken only as the more 



