On the Classification of the Diatomacese. By M. Paul Petit. 73 
outline with regular angles ; hut we never meet with the discoid 
form. The regular polygonal form affected by certain species of 
Triceratium * reaches to the limit of the discoid form characteristic 
of the tribe of Coscinodisceae. 
Genera, Isthmia, Biddulphia, Amphitetras, Triceratium. 
[Professor H. L. Smith unites the two last-named genera with 
Biddulphia, thus getting quit of the absurdity of calling a form 
with four or more horns a Triceratium. M. Petit’s remark that 
we never meet with a discoid form in this tribe is incorrect. Bid- 
dulphia radiata (Roper) is truly circular, and in a gathering made 
by Captain Perry, of Liverpool, Triceratium orbiculatum occurred 
with a circular outline and sub-marginal processes, in fact so like 
an Auliscus that many specimens were distributed with the MS. 
name of Auliscus formosus ; and it was only by the examination of 
a series of valves of T. orbiculatum from various localities that its 
true generic position was satisfactorily made out. On the other 
hand, one of the Coscinodiscese, (a species of Stictodiscus,) is some- 
times distinctly triangular. 
There is one characteristic which appears to be very constant 
in the genus Biddulphia, viz. two or more spines placed at right 
angles to the processes. These spines or setae may also be detected 
on several species of Triceratium, but usually placed at the base of 
the process. In this tribe, Professor H. L. Smith places Eucampia. 
-F. K] 
15th Tribe, Coscinodisceje (Fig. 27). 
In this tribe I comprehend all the genera having discoid valves 
and granular endochrome except Cyclotella. Meanwhile it is pro- 
bable that further study of the endochrome will result in a division 
of the tribe. Many of the species are only met with in a fossil state, 
and those that are living are for the most part found in great depths, 
which render the study of this group very difficult. The only 
species I have been able to study whilst living is Eupodiscus Argus, 
and the figure of Coscinodiscus centralis Ehr. by M. Max Schultze,f 
which seem to confirm the hypothesis that all other discoid forms 
have radiating granules of endochrome. M. Borscow 1 affirms that 
the Eupodisci have rudimentary appendages resembling the Bid- 
dulphias. 
Genera, Eupodiscus, Coscinodiscus, Actinoptychus, Asterom- 
phalus, &c. 
[I am not satisfied that Diatomaceae do five at great depths. I 
have had the opportunity of examining many deep-sea soundings, 
some of them from depths varying from 1189 to 3103 fathoms, and I 
have arrived at the conclusion that the forms occurring in them had 
* Kitton, ‘ M. M. J.,’ November 1874, p. 219. 
t Max Schultze, ‘ Q. M. J.,’ vol. vii., 1859, pi. ii. fig. 13. 
j Borscow, ‘ Die Susswasser Diatomaceen des Sudwestlichen Russland.’ 
