til") 2 SUMMARY OF CURRENT RESEARCHES RELATING TO 



may be of fixed composition (a pure metal, a definite compound, or a 

 eutectic mixture), or its composition may vary within a certain range 

 (a homogeneous solid solution). If the alloy is in a known condition of 

 equilibrium, reached by slow cooling or by quenching from a given tem- 

 perature, and the composition of the constituents is known, the propor- 

 tions of the metals present may be determined by measurement of the 

 areas of the constituents. It is often necessary to enlarge photo- 

 micrographs or to project them on to drawing or tracing paper. A 

 planimeter is used to measure the area of any particular constituent, 

 or if the pattern is complicated the drawing may be divided into 

 squares of 1 cm., and the proportion of one constituent, which may be 

 shaded in the drawing to distinguish it, estimated in each square. By 

 the planimetric method the composition of Muntz metal might be de- 

 termined with remarkable accuracy in half an hour ; this includes all 

 operations, from grinding to planimetric measurement. The method 

 failed to yield satisfactory values for copper-phosphorus alloys, the 

 explanation being the segregation of copper from the eutectic. The 

 correction to be applied was calculated, and the method then gave reliable 

 results. 



New Fatigue Test for Iron and Steel.* — T. E. Stanton has de- 

 vised a test which gives a combination of rolling abrasion and alternate 

 bending. A hollow ring of rectangular section, cut from the steel to 

 be tested, is placed between three hardened steel rollers. The upper 

 roller is loaded with a weight and rotated, thus imparting rotation to 

 the test-piece and the two lower rollers. The outer surface of the test- 

 ring is thus subject to rolling abrasion, and every radial section of the 

 ring is subject to alternate bending stresses which go through a com- 

 plete cycle three times in one revolution. A number of steel rails were 

 tested in this manner, at 800 reversals per minute. In the course of 

 the test the outer surface of the ring is worn down and spread over 

 the edges. In time small cracks appear parallel to the axis, and failure 

 takes place through the development of one of these cracks. The 

 number of reversals endured varied from 25,000 to 370,000. 



Metallurgical and Chemical Laboratories in the National 

 Physical Laboratory.'}' — W. Rosenhain, in the course of this paper, de- 

 scribes the rnetallographical outfit. The following details may be noted. 

 A Zeiss stereoscopic binocular Microscope is used for examination of 

 fractures. Two small rooms are provided for preparation of metal 

 sections, one is devoted to grinding (for which two carborundum wheels 

 are used), and emery rubbing, while the other is reserved for the last 

 stages of polishing and etching. A horizontal disk 9 in. diameter, 

 covered with cloth, is used for polishing. For etching steel sections 

 picric acid in alcohol and nitric acid in amyl-alcohol are employed. The 

 author describes his method of heating and quenching metal specimens 

 without contact with ah" : the metal is heated in an evacuated tube of 

 fused silica, through which a heavy stream of water may be directed when 

 the specimen is at the desired temperature. 



* Journ. Iron and Steel Inst., lxxvi. (1908) pp. 54-70 (9 figs.) 

 t Tom. cit., pp. 87- 108 (9 figs.). 



