The plate is next put through a rolling process so that all the grains are flattened out to form a mosaic covering over the whole surface. In spite of all the manufacturers can do there will still be some microscopic spaces between the particles, and these are filled up with a fine powder of carbon to prevent the passage of light. The screen is then coated with a very thin layer of varnish and upon this is laid a thin and extremely sensitive photographic emulsion. "And so that is the way these autochrome plates we have here were made," concluded the scientist. "Now our troubles begin, for we must be careful to give them a fair trial with the proper kind of an exposure and the proper kind of development." As the plates are extremely sensitive to all kinds of light the scientist cautioned the boy against loading the camera carelessly. It is better, he said, to load in a dark room. In putting the plates in the camera the plates are reversed and instead of placing the sensitized side toward the lens, the uncoated glass is put in front and the photograph is taken through the glass. Thus, the image first passes through the glass, next, through the grains of coloured starch, and, lastly, is recorded on the sensitive photographic emulsion. Before loading the camera, however, the scientist fitted a yellow colour screen over the lens, explaining that this was necessary to absorb some of the overactive blue-violet light rays, to which the emulsion is extremely sensitive. In exposing the plate what happens is this: Suppose a green field is to be photographed. The green rays of light, reflected from the field, pass through the lens, and through the glass support of the plate. But when they reach the coloured starch, the green rays pass through the green particles of starch, but not through the violet-blue particles, or the orange-red particles, for the grains of other colours absorb the green rays and hold them. Thus, development would show that the green light rays passing through the green starch particles caused the emulsion to darken under the green particles in just the proportion in which the green light reached them, and to record the image they carried. As the light would not pass through the other coloured particles they would not record any image. Thus a negative is produced, as we have seen, not the colour we see in life but the complement. By treating the plate with a solvent of silver the tiny black specks that were brought out behind each green particle are removed and each starch grain is allowed to transmit exactly the colour we see in life. In other words, we have a positive.