The main advantage of high voltage in electron microscopy is greater penetration. When using an aperture of optimum size the thickness of specimen that can be imaged increases almost linearly with applied voltage in the case of light elements, both when the criterion is image intensity and when it is resolution. For heavy elements the increase is less rapid. With a small aperture the increase in observable thickness is still less rapid, and ‘saturates’ towards I MV. For a specimen of given thickness, image definition increases nearly linearly with voltage owing to the decrease in chromatic aberration. Although ultimate resolving power improves with voltage, the gain is slight and is offset by a fall in contrast. The optimum voltage for very high resolution is probably between 200 and 300 kV. Radiation damage arising from ionization decreases with rising voltage, making easier the examination of sensitive materials such as polymers. On the other hand, ejection of atoms by head-on collision increases rapidly above a threshold voltage, causing observable defects in metals.

In construction, a high-voltage microscope differs from the normal type only in size and in having an accelerator instead of a simple electron gun. In operation it differs little, apart from precautions to avoid fiashover in the accelerator. A decrease in response of viewing screens and photographic emulsions is more than compensated by higher brightness of the electron gun. The chief applications so far of the high-voltage microscope have been for studying thick films of metals, magnetic materials, ceramics and polymers. Improved preparation techniques should make it possible to study sections of biological tissues up to 5 μ thick. The observation of micro-organisms and other specimens in the wet state can be carried out in double-walled cells, but only at poor resolution. Still higher voltages, up to 3 or MV coupled with the use of an energy analyser or an image intensifier, should improve further the microscopy of such thick specimens.

Learning and memory are important elements of our daily lives, familiar to all through introspection. Yet the mechanisms underlying these processes are still for the most part unknown. Here are problems which combine a maximum of intrinsic and practical interest with a minimum of actual knowledge and understanding. Years of our lives are dedicated to the formation of certain long-term memories and behaviour patterns, yet we have only rudimentary notions of how such ‘schooling’ is best accomplished. There is no certainty in any aspect of the process. We are not sure whether relatively few cells or millions participate in a memory trace; whether these cells change as a whole, or whether the changes are limited to synaptic regions. In fact, we cannot be certain whether the changes are confined to the neurones or whether the glia also participate.