2 9 8 D O C U M E N T 1 7 5 A P R I L 1 9 2 8 to give a sufficient deflection to a galvanometer needle. We will also suppose that the biological section is brought to a distance of about c.m from the particle. The arrangement is shown in the sketch below (which is not drawn to scale) We will also suppose—for the purpose of fixing our ideas, that the lower surface of the biological section consists of opaque strips each (say) c.m wide sep- arated by transparent intervals of the same thickness. If now we move the cover glass with the section attached over the colloidal particle, travelling transversely to the direction of the opaque strips and always preserving the same distance from the particle, it is clear that when an opaque strip is immediately above the particle, a less quantity of light reaches the photo-electric cell than when a transparent strip is immediately above the particle. And, if the galvanometer needle marks a travelling indicator sheet, the passage of each opaque strip will be recorded on the sheet, and in this way we have the germ of an instrument capable of a resolving power several times as great as that of the best microscopes.[14] To develop this idea into a practicable instrument, we need, in the first place, to devise means of moving the section, not in one direction only, but in two directions at right angles to one another, so that the all parts of the area of the section inves- tigated may be brought in turns over the particle and we have to aim at giving the section increments of motion of about c.m. The simplest means of achieving these small regular motions seemed to me to lie in having a differential screws made by joining inch and metric units. One could use two screws of pitches 1 m.m. and 1/25 inch or alternatively of pitches 1/2 m.m and 1/50 inch (25.4 m.m = 1 inch). In the former cases a single rotation of the screw head will move forward the sliding nut about .016 m.m. and in the latter case about .008 m.m. I consulted Hilger[15] a London firm who manufacture scien- tific instruments, and they said they could make a differential screw of this kind for, I think, £15. 10–6 [13] 310 –6 10–6