Cephalopd eye lecture - PLEASE HELP! :P


Pygmy Octopus
Hello - I'm preparing all the info I can find about Cephalopod eyes for a Neuphysiology class I'm in. The prof also teaches classes on sensory systems and doesn't know much about Cepholopod eyes, and has offered me extra credit if I bring him up to speed.

This is the update I sent him yesterday, asking him to ask me questions about the material. I'm sure he'd be more interested in info about the rabdomeres, rhabdomere convergence (what are the collateral axons all about there? Thats pretty similar to the human eye, and ) and is there more specific info about how the Ceph brain processes visual info? I couldn't find much more about the rhabdomere structure being split into 5 subreceptors, either, + that was interesting.


Octopuses have a slit-shaped rectangular pupil that changes in shape to accommodate light. In cuttlefish it is W-shaped, and in squid it is round. Nautilus don't have lenses at all, just pinholes open to seawater, but they still have retinas. Seems to work like pinhole cameras. Octopi, some squid have corneas, no other Cephalopods do. All other Cephalopods allow seawater into their eyes.

No apparent binocular vision in octopi, but cuttlefish do, for the normal reasons. One very interesting hypothesis I read was that the 'W' shaped iris in cuttlefish may form 2 images on each retina, making it double binocular vision. Octopi will bob up and down when looking at something. They also may be able to orient themselves in such a way as to allow both of their eyes to see the same thing (sorta like chameleons)

Cephalopods have a spherical lens, concentric with nearly spherical retinal surface. The lens telescopes out to adjust focus like a mechanical camera. Focal lengths are 2.1 to 2.6; Aperture about f 0.8. This is truly good performance. In theory, a refractive index of 1.67 (creating major 'fish-eye' distortion) is required to achieve this focal length, but this isn't the case. The lenses have been removed and tested, and they produce good images. After years of study, (decades), the solution was found to be to create a refractive gradient. (This would also have to change with size, and since squid grow (lenses over 40cm!), there must be a feedback mechanism ensuring the right gradient for the lens size.- still not understood) We still don't understand quite how it is achieved, but it seems to be achieved by varying the combination of crystallins. (crystallins - otherwise normal proteins pressed into service to be a lens) --- some interesting articles about the genetics of this:

Cephalopod photoreceptors are called rabdomeres. Usually the density is 20K - 100K rhabdomeres per square millimeter, but it gets upwards of 250K in deep sea dwellers. The rhabdomere is divided into 2 sections at the basement membrane. The topside directly faces the lens, and the underside of the basement membrane is divided in quarters - possibly producing 5 separate receptors? The topside uses rhodopsin (but different amino acid structure that vertebrates) The underside uses a secondary photopigment, retinochrome. Rabdomere support cells also have photopigments.

Cephalopods have a nerve going from each retinal photoreceptor to the brain, like insects. Cephalopod optic axons do give off collateral branches that, along with efferent fibers, forms a plexus beneath the retina. The efferent fibers seem to rearrange the photopigments.

There is no pit fovea. Some cephalopods seem to have an equatorial band that achieves the same purpose. (maybe this relates to the iris shape?)

Cephalopods seem to be mostly colorblind, although mostly this is from looking at the rhodopsin. Behavioral studies are inconclusive. However, they can detect polarized light quite well. Fish scales polarize light, thus making them easier to see by the cephalopod.

The flicker fusion rate is 20-60Hz.

Cephalopod brains are poorly understood, but the optic lobe takes up over half the size. The structure resembles vertebrate brains in form (striations, etc) and the optic nerve does make a spatially accurate map of the retina, and the subsequent layers underneath it do look very very similar to vertebrate visual processing centers.

Cephalopods can recognize shapes and categorize them. However, they have trouble with differentiating shapes that have been rotated (i.e. discriminating between a square and a diamond). This becomes almost impossible when the statocysts (vestibular organs) are removed ... and the image has to be carefully presented with respect to the rectangular iris. The information from the statocysts is used to keep the eye level, no matter what position the octopus is is.

Each cephalopod has two statocysts, surrounded by cartilage just outside the brain and behind the optic lobes. These are VERY silimar to vertebrate vestibules. They provide information on gravity, linear acceleration, 3 dimensional angular acceleration. At the end of every hair cell, is a nerve fiber that carries signals from the hair cells to the brain. Like the eyes, they are formed from an invagination of the ectoderm.

A slightly off topic note about chromatophores, that change the color of the Cephalopods for camouflage, mating, and other uses. Chromatophores change the density of pigments, and are operated by special muscles for speed, and I would presume to be voluntary. In vertebrates, chromatophores use microtubules to move pigments around (really slow!).


Pygmy Octopus
correction: The rhabdomere is not divided into 4. 4 rhabdomeres = 1 rhabdome, and is the functional unit. Also, the rhabdomeres at the center of the retina are longer and thinner than the ones on the periphery, as well as on the equatorial strip. There are many different sizes and shapes of rhabdomeres. The collateral axons forming the aforementioned plexus are efferent, and if they are removed then the rhabdomeres die, leaving the supporting cells (which also have photpigment). If the first layer of the loptic lbe is removed, same result. Removal of any otherlayer yields shorter rhabdomeres. However, the rhabdomeres interdigitate on their proximal ends, so they still talk to each other.


Colossal Squid
Staff member
Here are some notes for an article I've been putting together on ceph vision; rather scattered and incoherent, I fear, but maybe of some use:


cephs are often visual hunters
where did cephs evolve
compare/contrast deep sea vs shore (and even slightly amphibious)
convergent evo camera eye
biomorphic eng


from http://cephbase.utmb.edu/refdb/pdf/7329.pdf

Alloteuthis subulata 499nm
Loligo forbesi 494nm
Sepia Officinalis 492nm
Todarodes Pacificus 482nm
Paroctopus Defleini 480nm

from http://en.wikipedia.org/wiki/Rhodopsin

(see the graph of the 3 color curves and dashed rod curve)

human visual pigments:
red cone peak=564nm tail=680nm
green cone peak=534nm tail=650nm
blue cone peak=420nm tail=530nm
rod peak=498nm tail=600nm

so, assuming the octopus rhodopsin is about the same shape as human
rod rhodopsin but its peak is shifted to 480nm, the octopus' red end
perception will fall off at about 580nm.

As can be seen in this image ( from
http://en.wikipedia.org/wiki/Electromagnetic_spectrum )

this is sort of yellow to yellow-orange.

This means that an octopus can't see light frequency that's "redder"
than yellow-orange, but can see blues and greens just fine, probably
better than humans.

Note that although firefly squids are known to have 3 visual pigments,
and therefore color vision more like humans than the monochromatic
vision of most cephs, all three pigments are still in the blue-green
range: 470nm, 484nm, and 500nm. (Note, though, that the 484nm pigment
and the other two are located in different areas of the retina,
however) See http://www.springerlink.com/link.asp?id=p21210724627v321

most cephs can also see polarization of light, which humans
cannot. Some may also be able to see a bit into the UV range.

Kat recommends:

Young J.Z. 1960. Eyes, colours, and shapes. Proceedings of the
Royal Institute of Great Britain
. 38 (173): 401-413.

Eric Warrant & Dan Nielsen talk at Te Papa:


Eye size: Colossal and Giant squids have large eyes. Colossal may have
larger eyeballs, it's not clear which has larger pupils. Guess: deeper
needs low-light vision more.


vision drives crypsis
color matching w/o color vision
can't tell diagonals, but can tell horiz/vert
Robyn's visual navigation
didn't someone in Hanlon & Messenger show visual navigation in
O. vulgaris?
visual hunters, cuttles chasing shrimp
octos more tactile
vampyroteuthis sphincter eye
photophores, lens and reflectors
symbiotic bacteria
chromatophores, leucophores, iridiphores
extraocular photoreceptors
pupil shape
There's also a good chapter on ceph vision in Complex Worlds from Simpler Nervous Systems, Prete, ed. p.267-307

Some other good sources are Hanlon & Messenger Cephalopod Behavior, Nixon & Young The Brains and Lives of Cephalopods, and Wells Octopus. I've got a few other xeroxes of related stuff around... the most relevant seems to be from Symposia of the Zoological Society of London #38: The Biology of Cephalopods which has articles on "Pupilary Response of Cephalopods," "Extra-ocular photoreceptors in Cephalopods," and "Optic Glands and the Endocrinology of Reproduction" (since you mentioned being interested in this in another thread.)

Some other notes:

photopigments actually move around the receptors to adapt to light levels, so this augments pigmentation.

extraocular photoreceptors can detect ambient light levels for countershading

There are theories that the cuttle's W shape leads to two vertical slits intended to make sure that the light coming through the slits hits some specialized areas directly perpendicularly, rather than glancing.

Sorry this is a bit fragmented, but hopefully it'll help you a bit. Please feel encouraged to post what you find here, if you're willing. I'd love to see a list of your references; I don't think I'd seen the flicker fusion frequency listed anywhere before, for example, so I'm curious where that comes from...

good luck!


Pygmy Octopus
so, assuming the octopus rhodopsin is about the same shape as human rod rhodopsin
Its not the quite the same shape. It has a different amino acid sequence.

Whats the deal with the vampyroteuthis sphincter eye? I looked around a bit, but couldn't find anything about the eye... any links?

Whatabout the extraocular photoreceptors?


Colossal Squid
Staff member
lurker;128260 said:
Its not the quite the same shape. It has a different amino acid sequence.
From what I understand, the functional units of the opsins are fairly well conserved, though. But I didn't mean the shape of the molecule, I meant the shape of its response level graphed against the frequency of the light. And I'd love to have that actual graph, since GPO rhodopsin has been studied extensively, but all the papers I found just say where the peak is, not how wide the response is spectrally.

Whats the deal with the vampyroteuthis sphincter eye? I looked around a bit, but couldn't find anything about the eye... any links?

Whatabout the extraocular photoreceptors?
Sorry, no link for that one... just the references I had in the other post. I gotta run shortly, so I don't have time to re-read, but they are used for countershading or estimating the downward light levels in some species, maybe for daily migrations, and possibly for determining other environmental light-level stuff, maybe for seasonal spawning, etc.


Colossal Squid
Staff member
Any update on your presentation? I'm working on a ceph vision article for the TONMO articles section, and would be interested in what you found in your research.


Pygmy Octopus
yeah - I turned it in and got enough extra credit to bump my final neurophysiology grade to 100% :p (Never got 100% final grade in anything before, so I'm happy about that!)

What aspects are you interested in? There is a hot debate on intelligent design regarding ceph eyes vs vertebrate eyes. Seems that the inverted design actually generates too much heat and requires extra vascularization + energy. But ID folks argue the other way around, somehow.

Cephs eyes make a good model for gradual evolution, and eyes exist pretty much in the whole range from pinhole camera eyes (Nautilus) to the full meal deal, including lens, cornea, fovea, etc. (Octopus) Eyelids come in all shapes + sizes (even sphincters)

Some ceph lenses are stretched, like vert lenses, but others telescope out like glass lenses do. The refractive gradients are very precise and are established by varying proteins called crystallins (just like vert eyes.) Cephs have some mechanism to constantly maintain the accuracy of this gradient, unlike vertebrates, though.

Cephalopod light sensors are called rhabdomeres, and are arranged groups of 4 called rhabdomes. This is much like insect eyes. Also like insect eyes, Ceph eyes are designed to see polarized light. They can see vertical and horizontal eyes, but they are confused by diagonal lines. However, they can see every variety of polarized light accurately, even light that is circularly polarized.

Rhabdomeres send off axons or dentrites (the word I found was 'interdigitate' to each other, like human photoreceptors do. In humans, this works to be a contrast enhancer - photocells mutually inhibit neighbors, so only the ones with the strongest signal power through, and the ones that are 'close but no cigar' get inhibited. They call this signal processing in the eye itself.

Cuttlefish have a 'W' shaped iris that may present 2 images on each retina, presumably for extra depth perception accuracy.

The optic lobes of cephs take up about 50% of brain space.

Cephs brains are shaped as a donut surrounding the esophagus, meaning that they can only eat in small small bites or they'll give themselves a lobotomy. (no cephalopod snakes!)

lots more, but I can't remember it all. whats your email? I'll send you the powerpoints I made. Any more specific questions?


Colossal Squid
Staff member
thanks! And congrats on your 100%!

monty.mark@gmail.com would be great for the powerpoint slides.

Most of that is stuff I've got references on, but some of that I didn't know or had forgotten. I'd particularly be interested in references on the retinal "interdigitation" and maybe the lens rigidity vs. flexible bits. References are the ideal thing for me at this stage, since I'm trying to cite sources for everything, but write the article in an accessible style for the "educated layperson," to fit in with the science and fossils articles... I've got a big directory of papers I've downloaded, and a few books, and I'm trying to get it all organized into something coherent.

Anyway, I'm looking forward to reading your slides... I've been had this on a back burner for over a year, but between your interest and seeing a talk by Roger Hanlon a few weeks ago, I'm trying to get going on it for real (except that I'm going out of town for the holidays, and probably won't bring the big stack of cephalopod books....)


Staff member
OK Monty, we have a contest, your completing the eye research and getting the material on-line vs me finishing my Raising O. Mercatoris article. What's the prize?


Haliphron Atlanticus
dwhatley;130138 said:
OK Monty, we have a contest, your completing the eye research and getting the material on-line vs me finishing my Raising O. Mercatoris article. What's the prize?
Oh really... how's that article coming along D? :lol:


Staff member
gholland;130139 said:
Oh really... how's that article coming along D? :lol:
I have an outline, reread my original journal (I hope I write better than that now, or at least don't give octopuses tentacles) and an intro paragraph exposing my limited experience. Are you volunteering to be co-author?:sagrin:

the honor and respect of the masses?
You mean we don't already have that? :roll:


Haliphron Atlanticus
Doh! Co-author would be a just punishment since I also called them tentacles once upon a time! Actually, I'd be happy to assist in whatever way I can.... but you'd better get my time now before the spring semester starts up! :bonk:

Members online