188 SUMMARY OF CURRENT RESEARCHES RELATING TO 
steps in the evolution of this species : — (1) A small equilateral triangle 
is formed at the intercostal junction of the polygonal cells. (2) The 
angles of the triangle become blunted. (3) The blunted end becomes 
notched. (4) The notches deepen, and eventually becoming circular, 
form a perforation at each intercostal angle. At the same time the sides 
of the triangles form a larger perforation between them. These larger 
perforations are situated at the bisection of the sides of the hexagon. 
(5) The peripheral perforations situated at the intercostal angles break 
up into two, and, by repeated subdivisions, form the central finely per- 
forated membrane. (6) The peripheral perforations, when the central 
membrane is complete, become more or less of a uniform size, and then 
break up into tertiaries. (7) Repetitions of the 6th process produce a 
uniform and delicate perforated membrane over the whole of the valve. 
The 7th stage is conjectural only ; and, if the author’s view of the evo- 
lution is correct, the ordinary Asteromphalus form is but an intermediate 
stage in the full development of this diatom. 
Siliceous Valve of Diatoms. — Mr. F. J. Keeley, of Philadelphia,, 
sends us some diatom photographs, with a description of the mode of 
preparation, and the results obtained. 
After some rather unsatisfactory experiments in imbedding diatoms 
with the view of grinding sections similar to those prepared from natural 
cement-stones, the idea presented itself to Mr. Keeley that if he mounted 
broken valves with their edges in contact with the cover-glass and at 
right angles to it, and examined them with objective and condenser of 
large aperture, an optical section would result, of equal value to a true 
section, and having the advantage over the latter, if ground after imbed- 
ding, of certainty as to the identity of the form and freedom from inter- 
ference of the imbedding medium. This proved to be the case ; and 
some years ago he applied the method to a number of species, and com- 
municated the results obtained to the Academy of Natural Sciences of 
Philadelphia, but had neglected to write up the paper for publication.. 
The diatoms of which photographs were sent are Coscinodiscus Asterom- 
phalus Ehr., Actinoptychus heliopelta Gr., and Aidacodiscus Sollittianus 
Norm, from Nottingham, Md., and Triceratium grande Bright w. and 
Auliscus Oamaruensis Gr. and St. from Oamaru, N.Z. — species which 
show the typical structure of their respective genera to good advantage. 
The last species in particular shows a remarkable tubular structure in 
the comparatively thick plate of the valve. 
It is unnecessary to go into details as to conclusions, as the section- 
views obtained seem to admit of but one interpretation. In comparing 
other species examined with those included in the slide, Coscinodiscus 
Asteromphalus seems to fairly typify the genus, a middle honeycomb-like 
layer, with external plate containing symmetrical groups of apertures in 
each alveole and internal plate with but one aperture, having thickened 
margin, in each alveole, forming the “ eye-spot.” 
The Triceratia (or Biddulphise) of the favus type have a similar 
middle cellular layer, but the finer perforations are in the internal 
plate, and are arranged symmetrically with the centre of the valve. The 
outer plate contains holes almost as large as the “ honeycomb ” cells. 
Triceratium arcticum and the numerous species of similar character 
found at Sendai, closely resemble Coscinodiscus in structure, the finer 
