Abstract
It is proposed that the diatom raphe is a parallel-plate capillary containing a fluid which reacts at the trailing end, turning into a form which no longer "wets" the raphe walls, and which is left behind as a trail. More unreacted raphe fluid is drawn by capillary pressure from a source near the leading end of the raphe. This fluid sticks out from the raphe along its length, adhering to surfaces, thus causing gliding locomotion. Formulae are given for the maximum velocity and force of a moving diatom in terms of the raphe dimensions and the surface tension and viscosity of the fluid. An a priori estimate of the force exerted by a moving diatom, 1-50 millidynes, agrees with measured values. Five experimental tests of the theory are proposed.