A magnification of between 180 to 300 diameters would resolve all that was necessarv to be seen in most specimens, while in the case of cast irons even lower magnifications would serve. Sometimes a magnification of 15 diameters was of far more value than a photograph upon a scale of 2,000 diameters. It was, of course, necessary at times to use a high magnification for the purpose of proving a certain point."

The preparation and interpretation of the microscopic structures of metals was developed by Dr. Clifton Sorbv. His method was to cement to a plate a specimen 11/4 in. square and 1/16in- thick and then file it smooth ; the specimen was then ground with powdered stone and polished with rouge, finally being etched with 1 per cent, nitric acid in alcohol, washed and then dried.

Metals are opaque to transmitted light and must be illuminated by either obliquely or vertically reflected light for microscopic examination. With oblique illumination the light strikes the metal at an angle less than 90° and a smooth surface will appear dark and irregularities bright. This method is chiefly used for the detection of inclusions. When a beam of light is directed vertically on to the surface of the metal from a prism or a cover glass in the tube of the microscope a smooth surface will appear bright and irregularities dark.

The method of obtaining a specimen for microscopic examination is important, as shearing, oxy-acetylene cutting or breaking may change the structure. The best method is to cut out the specimen with a hack-saw. The specimen is prepared for examination by first grinding the surface flat, working down the ground surface on successively finer grades of emery paper and then polishing with an oxide (alumina). Soft metals are polished with some kind of metal polish and then freed from grease.

The etching reagents used to develop structures depend on the constituents to be identified. Normal structures in steel are developed with nitric and picric acids. Pure iron and pearlite are revealed if the carbon is below 0.89 per cent., but if above this, cementite is present and can be identified with sodium picrate. In the investigation of failures of steel it is important to ascertain the distribution of impurities, as if these are segregated their influence will be greater than if evenly distributed. For alloys containing two metals and solid solutions different methods must be employed and the crystal structure developed. With large crystals of lead it is interesting to note that a Brinell test does not show a circular impression. Pure metals which have been worked cold and annealed often show twinned crystals. In some alloys the two constituents are not dissolved in each other. Steel develops a similar structure by the separation of ferrite or cementite from a solid solution. A steel is composed of ferrite and eutectic ; as the carbon increases the eutectic increases until at 0-89 per cent, of carbon, the whole structure is pearlite ; above this point is carbide of iron and eutectic. Lastly, there is the " solid solution " type of alloy. A brass composed of 70 per cent, copper and 30 per cent, zinc solidifies in much the same manner as the copper and nickel alloys, producing " coring." If the alloy is worked cold and annealed the coring disappears and the structure becomes twinned.-(The Photomicrographs Society.)

Fig. 157 illustrates a complete apparatus for the taking of photomicrographs, this particular apparatus has a total length of 6 ft. 3 ins., a width of 101/4 ins., and a height to centre of 10 ins. The camera takes plates 61/2 x 43/4 (1/2-plate size) and the bellows extends to 36 ins. The front portion of the base carrying the illuminating apparatus and microscope may be swung out at right angles to facilitate visual focussing.