THE MICROSCOPE. 133 
forms of Campylodiscus, Surirella, etc., and that, so far from upset- 
ting it, tended to confirm former theories as to the structure of the 
Diatomacez. 
In the list of genera reported by Mr. Christian (loc. cited) as 
found in the same deposit, Campylodiscus and Surirella do not 
appear; but with Actinocyclus, Auliscus, Eupodiscus, etc., which 
are reported, Campylodiscus usually keeps company, and may 
reasonably be looked for, where these are found. Species of Campy- 
lodiscus are found having a cellulate marginal rim, a naviculoid 
centre, a meridian line, and radiate moniliform striz prolonged to 
the margin of the disk (C. singalaris, C. panduriger Schm., C. 
exilis, Grunow, ete.) These have not the meridian line “inter- 
rupted,” but this characteristic is found in C. Greffii, Grun. None 
of the foregoing species have “nodules” at the ends of the meridian 
line, such as are shown in the figures of Raphidodiscus but not 
mentioned in the description. Such nodules are found, however, in 
C. imperialis, Grev. and C. ecclesianus; ibid, in many of their 
varieties. 
In the genus Swrirella all of the generic characters of Raphi- 
dodiscus may be found, even the discoid form (8. Crumena Breb.), 
but nowhere combined in the same diatom. Auliscus, although having 
in some cases a meridian line and naviculoid centre, has not a cellu- 
late marginal rim. 
In the case of fossil diatoms our only basis for judgment as to 
the classification is the structure of the frustule or valve, its habitat 
and mode of growth being usually merely conjectural. Guided by 
the structure of Raphidodiscus, what are its affinities and where 
should it be placed among the tribes, families and genera of diatoms, 
following the classification of Prof. H. L. Smith, by whom its generic 
name was assigned? The original name given to this diatom by 
Mr. Peticolas, Navicula disciformis, indicates that he regarded its 
structure as that of the typical Navicula, a flat valve with a central 
nodule and a true raphe, symmetrically dividing the valve and ter- 
minating in a nodule at or near each end of the valve, the specific 
name relating merely to its contour. The significance of the generic 
name, Melonavicu/a, under which this diatom next appeared, is not 
clearly apparent; and, whether derived from Greek or Latin root, 
was founded on unstable characteristics and was ill chosen, and, for 
that reason, probably was abandoned. 
The latest generic name, Raphidodiscus, makes the discoid 
character of the valve and the presence of a true raphe the basis of 
nomenclature, and being based on structural characteristics which, 
