This branch of science is of increasing interest and importance, and the following extracts from a paper read by Mr. G. Mohn, M.Sc, at a meeting of the Manchester Microscopical Society give a very clear description of the necessary procedure and methods required to carry out the desired study.

" The science of Metallography, as the study of metals through the microscope was called, was practically a new subject. It was perhaps only during the past 20 years or so that the microscope had played any real part in metallurgy, but in that short time the science had extended to such an extent that the student could easily devote a lifetime in studying one small section of one alloy system. He did not intend to deal with the actual science of metallography, but to show the practical way of bringing specimens of metals into such a form that their structure could be examined by means of the microscope, and how these structures could be photographed.

A metallurgical specimen was rather different from a petrological specimen in that it was composed of a metallic body which was opaque and transmitted light could not be used. In the first place, the prepared specimen should show the part which it was necessary to examine. If it was desired to examine into the cause of the fracture of a wheel, for instance, a specimen should be prepared from the fractured region as well as from the undamaged portion of the wheel, in order that the two might be compared.

An ordinary specimen looked at through the microscope, just as sawn off, would not show anything except at a very low magnification ; at a high magnification the surface would be too rough for the microscope to be any use. The surface was therefore, polished so that it became,perfectly level, quite bright, and free from scratches. The specimen was first ground on an emery wheel until flat, and then polished by different grades of emery paper. The first emery paper used was of about the same coarseness as that in household use, and then finer and finer grades, until, eventually, a paper was used of which the abrasive quality was hardly perceptible even by rubbing a finger along it. Having arrived at this stage the specimen, to the unaided eye, would appear quite smooth and flat, yet, when examined under the microscope, would present an appearance similar to that of a ploughed field. Further polishing must be done upon a polishing wheel of Selvyt or lamb's skin, using slight abrasives. The abrasives which were generally used for steel were diamantine (ignited alum), rouge, or chromium sesquioxide, very finely ground, and levigated, so as to be free from all gritty materials. For such materials as copper alloys, brasses, and bronzes, " Brasso " was an excellent polishing medium.

After all this preparation, the specimen should be absolutely free from scratches, and its surface present the appearance of a polished mirror. When examined under the microscope every constituent in it should appear equally polished and ground to the same level, and the field should appear to be brightly illuminated. Any mechanical inclusion in the metal would then be clearly shown ; i.e., such as slag in wrought iron, which was included in the rolling processes, or graphite in cast iron, graphite being the pure carbon which separated out on solidification and during the cooling of the cast iron.

The method adopted for making the structure readily discernible was the etching process, which was a sort of differential attack on the surface of the specimen. There were many etching reagents and they differed considerably in their action. Picric acid was commonly used for steel and acid ferric chloride for brasses. It was quite probable that some of the constituents present on the surface to be etched were electro-positive to others and that these were attacked and discoloured more than, or to the exclusion of, the electronegative constituents, a differential effect being thus obtained.

The fact that even pure metals could be etched might seem rather strange, because pure metals were perfectly homogeneous. The next illustration was one of pure gold, in which some of the crystal grains appeared dark and others light, while around each grain there was a definite line. The reason for this was that although there was no question of electro-positive and electro-negalive constituents, the orientation in each grain was the same throughout, but the orientation differed from grain to grain. Students of crystallography were aware that the rates of attack were different in different directions in a crystal, and grains with a certain orientation would be acted upon at a different rate from other grains where the orientation was different. The grains, therefore, showed up light or dark according to the differential corrosion caused by the difference in orientation on the surface.

A solid solution of an alloy was then shown in which the material should have been homogeneous. Owing to the fact that the cooling had probably been fairly rapid there was a certain amount of difference in composition between the first crystals to solidify and the last, the difference in composition being quite sufficient to show a variation between the inner part of each crystal and the outer part, so that the etching had been differential, due to the difference in composition. Such a specimen, on annealing, would be similar to a specimen of pure gold.

Etching reagents attacked different metals in different ways. For example, in a specimen of high carbon steel containing " Pearlite " (iron and Fe3C in intimate mixture) and free Fe3C ; the iron carbide Fe3C showed up perfectly white when etched with picric acid. Boiling sodium picrate etched the Fe3C black, and often when there was doubt as to whether the excess constituent in a steel was Fe or Fe3C, etching with sodium picrate would solve the difficulty.