From: Ben Craven To: 'Cathy Colwell' ; 'lempel@n-jcenter.com' Subject: RE: Turtle Project At Mainland High School Date: Tuesday, October 27, 1998 8:25 AM Dear Cathy and Brian, Thanks for both of your e-mails. My understanding of your problem is that you'd like to find out if it's possible to have lights that are acceptable to humans that are invisible to the turtles. Out of all of the electromagnetic spectrum, from long-wave radio to gamma rays, there is a tiny "window" of wavelengths between 400-700 nm which are "visible light" as far as humans are concerned. This window arises because of the cones (photosensitive cells) in our retinas. There are three kinds of cone. Each is sensitive to a different range of the spectrum (the ranges overlap) and between them they cover 400-700 nm. Similarly, there will be a window in the e-m spectrum which is visible light as far as turtles are concerned. The position and width of this window depends upon the number of types of the turtles' photoreceptors and their spectral characteristics. It is quite possible that the turtle has only a single kind of photoreceptor (most nocturnal creatures are like this). How does this knowledge provide strategies for designing lights? Well, as you suggest, the thing to do is to try to use lights to which you know the turtles are insensitive, ie which contain wavelengths outisde their window of visible light. Now the only way to do this is to find out exactly what the spectral characterstics of the turtles' photoreceptor(s) are. Doing this using "psychological" methods on any creature that can't speak is very very laborious, and I'd be very surprised if it's been done on turtles. However, if the past 20 years or so physiological methods have been developed to do the same job, and maybe someone has applied them to turtles. To find out, you need to search one of the scientific databases, using keywords like "cone", "photopigment", "spectral sensitivity", "microspectrophotometry" and of course "turtle". I'd have a go myself but our access to databases has just been reorganised here and I haven't learned the new system yet. If you find anything that looks hopeful, but can't follow it, send me the abstract and I'll have a look. Assuming you are successful, there are various possibilities. The worst is that you find that the spectral windows of humans and turtles overlap greatly. If this is the case, then lights will have to be restricted to deep reds. If the overlap is less, then maybe reds and oranges might be permissible. An intriguing possibility is that if the turtles has only a single type of photoreceptor, their spectral window will probably be a good deal narrower than ours, and it may be that they are insensitive to both red and blue, in which case deep red, deep blue, and purple lights might be acceptable. Anyway, all of this is guesswork unless we can track down some facts about the turtles. What is certain is that if we only have a small region of the spectrum to construct lights with, we can only produce a very restricted ragne of colours, and only with at least 2 widely separated wavelengths have we any chance of making white. I hope that this is of some use. I attach the handout of the lecture that, funnily enough, I am giving this afternoon on the topic of the varieties of colour vision. If only you could come along! The second attachment concerns the physiology. Good luck and keep asking! Ben <> > <> ---------- >