144 
PROCEEDINGS OF THE SOCIETY. 
mediate layer ; a layer that I certainly have never been able to demon- 
strate in this genus except in P. formosuvn , and what is more to the 
point, does not seem to be demonstrated in this photograph. 
To me the weak point in all these photographs is that there is 
nothing to show the difference in appearance between the two sides of 
the valve ; the first four from the w r ant of sufficient aperture in the 
lenses themselves, and the last two because the increased aperture has 
not been allowed to act owing to the stopping down of the sub-stage 
condenser. All the pictures are true enough as far as they go; the 
outer side of the valve, with the widest aperture of the lens, does give 
this appearance of white hexagons with dark interspaces, but when you 
examine the other side of the valve with the same aperture the appear- 
ances are exactly reversed, and you have dark hexagons with white 
interspaces. I wish it to be understood that I am using the word 
hexagons in a conventional sense only, for it is a question with me 
whether they have any real existence on any side of the valve. I have 
no doubt in my own mind that the structure on the outer side of the 
valve of P. angulatum is of exactly the same character as that on the 
corresponding side of P. formosum , and likewise consists of short bars 
of silex set lengthways on the valve as the following examples will 
show. 
The inner side of the valve seems to offer the greater difficulty of the 
two, owing to its being the more robust, and showing but few examples 
of torn structure ; but lately I have found what may be considered a 
unit of structure for this side which should help us a little towards 
understanding its true meaning. To procure an appearance of dark 
fields with white interspaces when the inner side is uppermost, is purely 
a result of wide aperture, when the valve is mounted dry on the cover ; 
but the shape of the fields will depend on how much of that aperture is 
utilized. With a small aperture of the condenser they conform to the 
appearance of the other side of the valve and are white ; but when the 
aperture is enlarged sufficiently they assume the shape of dark rect- 
angles. With full aperture of the dry condenser and bull’s-eye they 
become hexagonal, and photographically the boundaries of the hexagons 
are often split up into six intercostals. I say photographically, for 
visually intercostals are not in evidence, or at least not to me, having 
spent two whole evenings trying to see them without doubt. Of course 
I have seen them, but every microscopist knows what that means when 
looking for some minute detail which he knows ought to be there, and 
imagination takes the place of actual vision. 
The fact of not being able to see them ought not, however, to be 
conclusive evidence against their existence, when we consider within how 
minute a space all these subdivisions are contained — the whole structure 
in a blaze of light — and it is easy to suppose that the photographic plate 
is more sensitive to such minute differences in illumination than the eye 
would be. Still I am awake to the danger of trusting to appearances 
only when minute details are concerned, and remember that Mr. Nelson 
has shown us how in many cases intercostals are the result of the 
secondary spectra and that a wide-angled objective becomes a small one 
in relation to very minute structure. But while it may not be possible 
to determine from the appearances alone that the intercostals are real, 
