The Three Mirror Telescope (3MT) has been developed at the Institute of Astronomy by Dr. Roderick Willstrop. Its optical design is unique because it is the only form of telescope which combines the three advantages of a wide field of view, very small sharp images, and all-reflection optics.

Reflecting telescopes have been built with apertures up to 10 metres, but they have fields of view limited to 40 arc minutes (Ritchey- Chretien two-mirror design) or 1 or 2 degrees with a 3- or 4-lens Wynne corrector near to the focus of the main mirror. Schmidt cameras can give fields of view of 7 or 8 degrees, but their apertures are limited to about 1.3 metres because they use a thin glass lens large enough to cover the whole aperture.

The 3MT has a field of view 5 degrees in diameter, and the ray- theoretical image size is less than 0.33 arc seconds everywhere, and less than 0.1 arc seconds over the central 1 degree of the field. Because no lenses are needed the images are perfectly achromatic, and furthermore it is possible, in principle, to build this telescope with a larger aperture than any Schmidt camera.

The design is based on a simpler one discovered in 1935 by the French optician Maurice Paul and rediscovered independently in 1945 by the American James Baker. The original Paul-Baker design had a paraboloidal primary mirror, a convex spherical secondary mirror and a concave spherical third mirror. If the aperture of the primary mirror was f/4, it would have given acceptable images, and a field of view of about a degree.

The good performance can be explained as follows: if the second mirror were a convex paraboloid (instead of spherical) the light from a distant star would be made parallel again after the second reflection. (This arrangement of two coaxial and confocal paraboloidal mirrors was described by Mersenne in 1636.) The third mirror would then also have to be paraboloidal to focus the light, and the field of view of the whole system would be no larger than that of a single paraboloidal mirror. The essential feature of the Paul-Baker design is that the second mirror is spherical, so the light is not exactly parallel after the first two reflections, but is deviated in just the same way as by the corrector lens of a Schmidt camera. Then the third mirror must also be spherical to focus the light, and a large field of view with sharp images is obtained. This telescope has also been called the Mersenne-Schmidt.

Baker suggested that the Hale 200-inch (5.08 metre) telescope might be given a wider field by using two auxiliary mirrors. It would not have been acceptable to enlarge the central hole in the irreplaceable 200-inch primary mirror. The third mirror, 2.1 metres (84 inches) in diameter and weighing one or two tons, would therefore have been mounted directly above it, so the system was never built. The field of view could have been just over 1 degree. In the f/3.3 Hale telescope, it would have been necessary to modify the shape of the secondary mirror to retain small, sharp images.

The field of view has been increased here to 5 degrees by placing the third mirror behind the primary, and by increasing the relative aperture from f/3.3 (in the 200-inch proposal) to f/1.6. To retain excellent image quality with this large aperture and field it was also necessary to make small changes in the shapes of all three mirrors from the paraboloidal and spherical shapes that were satisfactory in the original Paul-Baker design.

Two versions of the 3MT have been built; a working model of 100 mm (4 inches) aperture (completed in 1985) has been used to take some photographs of the sky, and a prototype of 0.5 metres (20 inches) aperture (1989) has been used both for photography and with a CCD, and to test methods of aligning the mirrors so as to obtain the clearest images. The sky in Cambridge is nowadays too bright for front-rank research on faint extended objects with such a fast camera as these. The justification for building the prototype was to demonstrate that the design works as well as the Schmidt camera, and so to lead to the building of larger cameras of this type on much darker sites.

The prototype 3MT is on an equatorial mounting which accounts for nearly half of its height of 2.5 metres. The focal length is 800 mm, and the tube length 1.2 metres, which compares very favourably with the Schmidt camera in a neighbouring dome: this has a tube more than 4 metres in length, although its light-collecting area is very little larger than that of the 3MT. The 0.5 metre 3MT stands in a simple wooden building about 3.6 metres (12 feet) square, with a roof that runs off on rails to the North. It has been provided with a control system (see link above).

A proposal, by a committee of the Royal Astronomical Society in 1986, to build a telescope of this type with an aperture of 5 metres was not pursued after a committee of the Science and Engineering Research Council in 1987 recommended that UK funds should be used to purchase a one-quarter share of the 8-metre Gemini telescopes. However, a proposal by J.R.P. Angel of the University of Arizona to build a telescope of this type of 8 metres aperture, slightly modified to allow the use of many CCDs, was approved by the U.S. National Science Foundation in 2000.

Star field, 30' x 38', in Cygnus, centred at R.A. 21h 50.7m, Dec. +52 11'. Recorded using the 3MT on 1997 Sept. 26, 120s exposure, Hour Angle 20m W at mid-exposure, with a Kodak KAF 1400 CCD behind a Schott RG630 filter. The field of the telescope is ten times the width of this frame. North is up.