When University of Idaho researchers contemplated designing a 300-m, open-path Fourier transform infrared (FTIR) spectrometer, they realized that the mirrors they needed would be heavy, would vary with temperature and would cost a lot. So they decided to build their own. Peter R. Griffiths, a professor at the university, and Robert L. Richardson, then a postdoctoral associate, selected epoxies based on viscosity, rate of polymerization, shrinkage, strength and coefficient of thermal expansion. They mixed particulate graphite with the epoxy and poured the fluid onto a spinning blank. They poured multiple layers of epoxy, each successively thinner and less viscous. With some variations in the process, the researchers made composite surfaces that varied from 40 cm, f/0.5 to 76 cm, f/>100. When cured, no surface grinding or polishing was necessary. The researchers coated the surfaces with a commercial silver spray gun. The mirrors are fairly soft, so they are easily scratched, but they are five to 10 times lighter than their glass equivalents. They also are strong. "We actually dropped a cats-eye retroreflector from a height of about 5 feet," Griffiths said, "and it survived perfectly intact with no loss in reflected signal -- something that would never happen with glass." As for cost, he said that they spent $100 to $200 in materials constructing a 16-in. mirror. These are not production items, however, and Richardson, who helped develop the process and co-authored a paper reporting the work in the February issue of Optical Engineering, has moved on to Command Technologies Inc. of Satellite Beach, Fla. "Frankly, I would love to use mirrors like this in our current open-path FTIR investigations," Griffiths said. "However, it needs somebody who really wants to make them.