New Diatoms from Honduras. Described by A. Grunoiv. 183 
[I. capensis is perhaps the I. minima of Harvey and Bailey. — 
F. K.] 
Triceratium. 
T. ( orbiculatum Shadl. var.) elongatum Grun. T. frustulis 
cyhndraceis concatenatis, membrana connectiva elongata lineis 
transversis (e divisione imperfecta ortes ?) plus minus numerosis in 
utroque dimidio opposite curvatis instructa, yalvis exacte orbicularis, 
vel in tribus locis pluries undulatis, varius late ovalibus, cornibus 
tribus (rarissime quaternis) conicis obtusis aculeo unico instructes, 
nodubs suborbicularibus, marginatis ab margine aliquantulum re- 
motes tenuissime punctatis. Areolae bexagonae parvae in valva 
radiantes in membrana connectiva lineas longitudinales et obliquas 
efficientes. Diameter valvae ■036-' 102. Tab. CXCVI., Fig. 2, 
a, b, c. 
Common in the Honduras gathering. 
This Triceratium is distinguished by its orbicular valves and 
the slight undulation which I have not observed in other specimens 
of T. orbiculatum. It closely resembles the elongated form of 
Auliscus pruinosus as figured by Bailey, and also a peculiar 
Cerataulus which I shall presently describe. 
Cerataulus. 
I have in my collection a form of Cerataulus from China (legit 
Gaudichaud) which I consider to he a dubious variety of C. lievis, 
and have named it 
C. (lievis var. ?) Chinensis Grun. It differs from C. lievis by 
its truly circular valves, the greater number of processes, and the 
arrangement of the minute dots which are irregularly scattered over 
the whole surface, and only form short striae (13-14 in ‘01 mm.) 
near the margin of the valve when seen from above. The con- 
necting membrane is sometimes elongated (PI. CXCVI., Fig. 3, a, b) 
like that of Triceratium orbiculatum var. elongatum, and shows 
similar divisions separated by curved lines, and which are always 
constant in the connecting membrane of Rhizosolenia. 
I am unable to determine whether this elongation of the con- 
necting membrane is connected with some at present unknown 
method of propagation, or only the result of certain abnormal con- 
ditions of life. The frustules of Orthosira Roeseana (Babenhorst) 
( = 0. spinosa Greville) often become elongated, and have then a 
very complicated connecting membrane. These abnormal forms of 
0. Roeseana seem to be produced in places where they are not 
always covered with water. Ehrenberg found them on the stems 
of trees, and named them Liparogyra spiralis and L. dentroteres. 
I have also found them on the thallus of a Marchantia from China, 
and sometimes in gatherings of 0. Roeseana, but less developed. 
