D O C U M E N T 1 7 5 A P R I L 1 9 2 8 2 9 9 Either of these screws could enable regular motions of the order of cm to be made. So that this part of the problem seemed to be solved. I am of course supposing that, as different parts of the section were brought, in order, over the colloidal particle, so a picture is formed as in ordinary tele- photography by a spot of light whose intensity depends on the amplified current from the P.E. cell travelling in a similar order over a photographic plate.[16] If ABCD represents the area of the section to be investigated, one would proceed from A to B, bringing each of the ten squares in the vertical column successively over the particle, then advance to the next column and come back in reverse direc- tion to the AD line. One would however aim at investigating an area consisting of to of such elementary squares. Both motions i.e of the section & the photographic plate would of course be pro- duced by synchronised clockwork. The next problem concerns the means of making a biological section which should have a truly plane surface over the small area investigated. I think that I solved this problem also, but since the description of the method is a little long, I will pass over it here.[17] The method would also serve—as seems very neces- sary—to make sections much thinner than .001mm. Then there is the question—how can the section be brought sufficiently close to the colloidal particle without the risk of touching it and displacing it in the course of the adjustment? This seemed to be a difficulty so long as the particle was simply deposited on top of the quartz slide. And it appeared to me that the glass ought to be manufactured with the particle in it, and then the slide or plate carefully ground down until the particle were exactly in its surface. It is, of course, not necessary that the surface of the plate should be truly plane. It would be sufficient that the particle should lie in the highest part of it. It would answer well if the plate were of shape as below (not drawn to scale) I have no doubt such a plate could be prepared without any difficulty by one of the glass manufacturers. 10–6 502 1002 [18]