ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 221 



identifying an alloy of two constituents of different colours by citing 

 an alloy of 54 p.c. of gold and 4G p.c. of aluminium. This contains a com- 

 pound of a beautiful purple colour, discovered by Prof. Roberts- Austen, 

 and having the definite composition of AuAl 2 . It possesses the very 

 rare property of having a melting-point superior to the melting-point of 

 the less fusible of the two constituents. During solidification it separates 

 first as crystallites, which appear black in the photograph (plate II. 

 fig. 1) ; the interval being filled by a matter of subsequent solidification, 

 whose proportion and nature vary with the composition of the alloy. 



Plate II. fig. 2 illustrates the process of testing by abrasion upon a 

 soft foundation by the assistance of very fine powders. The sample was 

 a hard steel bar (1*24 p.c. of carbon), forged into a round bar, 12 mm. 

 in diameter, the forging having ended at a dark red. A cross section 

 was polished, and the polishing continued upon apiece of wet parchment 

 covered with a very small quantity of rouge. The iron became dug out, 

 placing in relief the carbide Fe 3 C, which in this sample is divided into 

 a multitude of small rounded grains, which appear dark upon a light 

 background, or light upon a dark background, according to the position 

 of the objective. 



The same method is convenient for resolving clearly the mixture of 

 iron with the carbide Fe,C which was discovered by Dr. Sorby to be a 

 constituent of all slowly-cooled steels. The mixture is in layers, usually 

 curved, the thickness of a ipair of lamellre averaging about O'OOl mm. 

 From the irisated effect of the lamellae the name of pearlyte is given 

 to the structure. 



If the fusibility curve of a two-metal alloy (such as copper and 

 silver) be studied, it is found to contain two descending branches 

 crossing off at a certain temperature (770° C), where they are simul- 

 taneously crossed by a horizontal branch. Thus at this temperature 

 there is a triple point. The explanation appears to be that above the 

 triple point the metal in excess separates out, while below the eutectic 

 alloy (i.e. the alloy in which the metals are mutually saturated) separates. 

 Inasmuch as Leval found that this alloy had a definite composition 

 Ag 3 0u 2 , and as it crystallises in a distinct form, it was long thought to 

 be a chemical compound. But when reheated it is not found to be 

 homogeneous. Some of it (Cu) takes an orange coloration, and some 

 (Ag) remains white. The orange metal is distributed as small dots or 

 as plates (straight or curved) alternating with the white plates of silver. 

 (Plate III. figs. 1, 2, ; plate IV. fig. 1.) These alternations moreover 

 produce a pearlyte-like effect. The evidence therefore seems to show 

 that the eutectic is a mechanical mixture and not a chemical compound ; 

 although it is singular that its formation never begins until the unsolidi- 

 fied mass has reached the definite composition of Ag 72 p.c. and 

 Cu 28 p.c. 



Slowly cooled steel shows a similar fusibility curve, but the branches 

 above the eutectic point correspond to the separation of pure iron 

 (ferrite), or of Fe 3 C (cementite), whichever is in excess, and the hori- 

 zontal branch to a simultaneous separation of the two constituents in 

 alternating layers. This eutectic point (670° C.) is sometimes called the 

 point of recalescence, as the diminution of volume is interrupted and a 

 sudden evolution of heat takes place, sometimes evident to the naked eye. 



