of the Solar Spectrum on the Daguerreotype Plate. 129 
another unexplained difficulty. While on the one hand it is 
obvious that the dispersed colours arise in a totally different 
manner from the specular ones, and cannot be regarded as 
directly connected with them; on the other it does not ap- 
pear by what superficial structure, by what form and law of 
distribution of laminze adhering to it, or by what hypothesis 
respecting their crystallization, &c., any such colours can be 
explained. One thing, I think, seems clear, that they cannot 
be produced by any lamination parallel to the surface of the 
plate ; and another, that the laminz (if any) which do pro- 
duce them must be disposed indifferently in all azimuths, 
and (within certain limits) at all angles of inclination to the 
general surface, since they are seen in all directions. 
The absence of the diluting white light which mingles with 
the pure tints in Newton’s transmitted rings, is also a very 
peculiar character, and would almost incline us to a suspicion 
of the origin of these dispersed colours being to be sought 
within the domain of double refraction and polarization, were 
it not extremely difficult to conceive any structure in which 
such a mode of production can be realized. 
It is not improbable that this obscure part of the subject 
will receive elucidation from the halo-colours which appear 
on the white ground, and on the white portions of the spec- 
trum when illuminated by rays nearly approximating to the 
direction of specular reflexion. In such incidences only 
they appear, and then only at the borders of the ground 
where the action of dispersed light has been feeble, and along 
the region A and the faint train G, which, when the illumi- 
nating light is remote from specular incidence, are very faintly 
illuminated, but, as it approaches that incidence, seem to open 
out as it were from B and D in brighter light, coloured with 
faint pink. In the same manner the frosted ground under 
these circumstances dilates in breadth, so as to cover a much 
larger portion of the whole plate. Its borders gleam with the 
pink tint, while its internal portion passes into dark gray, 
which increases up to the specular incidence; the order of 
tints in this situation being precisely the first order of colours 
of the Lunar corona, or of the halos exhibited by a candle 
seen through a glass breathed on. 
These observations may seem unnecessarily minute, but I 
am persuaded that they are essential to the right understand- 
ing of the Daguerreotype phenomenon, when we shall arrive 
at one. Already we see in our reference of the phases of this 
phznomenon to the colours of thin plates (with whatever dif- 
ficulties that reference may be encumbered) a very simple 
explanation of the change from positive pictures to negative, 
Phil. Mag. S, 3. Vol. 22, No. 143. Feb. 1843. K 
