84 
GrevillEj on New Diatoms. 
In general aspect, this whimsical-looking diatom comes 
nearest to S. Ameincanum, but is more minute, has a smooth 
surface, and the capitate valve is furnished with a spine or 
bristle between the processes. Thefrustule is subject to con- 
siderable variations, which may probably be accounted for by 
its progress toward the period of self- division. In its early 
stage the diatom is filamentous, specimens in my possession 
showing three frustules in situ (Fig. 22) ; and it will be per- 
ceived that the capitate ends of the frustules are opposed to 
each other within the tube, the horns meeting, and the 
terminal spines overlapping each other, as in the genus 
Hemiaulus. It appears that the frustules become fully 
developed within the tubes before they escape, for although, 
in the figure just referred to, the conical process is not yet 
visible, it is mature, or nearly so, in Fig. 24. When perfect, 
the frustule may be described as comprised of two valves, 
rather sharply quadrangular (as viewed in the microscope) ; 
the one suddenly contracted and passing into a long conical 
process terminating in a minute spine j the other contracted 
into a short, thick neck, supporting a spherical head furnished 
with two short, conical, truncate horns, each tipped with a 
spine at its inner angle, while a slender, short spine is also 
situated in the intermediate space. The relative proportion 
of the two valves varies greatly, as will be seen by consulting 
the figures; but this seems to be of no momemt in a diagnostic 
point of view. A very remarkable deviation from the typical 
structure of the genus occurs in S. simplex, Bail."^ A very 
minute species, in which the valves are described as ^' nearly 
symmetrical,''^ and ^' both gradually tapering into pyramidal 
cones.^'' 
It is to Mr. Palmer that we are indebted for specimens 
in so perfect a state as to indicate distinctly their affinity 
with Hemiaulus. Nevertheless, we cannot but agree with 
Mr. Kalfs in his observation, that, ^' although it is not diffi- 
cult to point out differences between the Chcetocerece and other 
groups, yet, on account of the variety in their forms, we con- 
fess our inability, in the present state of our knowledge, to 
give a concise definition which shall include its own members 
and exclude all others.^^f 
Navicula. 
Navicula spectalissima, n. sp., Grev. — Elongated, deeply 
" Notes on new species of microscopical organisms, chiefly from tlie Para 
Kiver, South America/' By Loring W. Bailey. ' Boston Journal of Natural 
History,' vol. vii, p. 343, fig. 65. 
t Pritchard, * History of Infusoria,' 4th edition, p. 860. 
