Manchester Memoirs, Vol. li. (1907), No. 15. 7 



they shoot out arms, seizing territory that they are after- 

 wards going fully to occupy. These arms come out and 

 meet and interlock, and then the places between are filled 

 by a slower growth. The ultimate effect is to produce a 

 solid block in which the skeleton formed by the original 

 shooting out of the arms is completely merged. You 

 may see the same kind of thing happen when a salt 

 crystallises out of a solution on a glass plate. The 

 different crystals are liable to become interlocked at their 

 junctions by means of the arms shot out from each. It 

 appears (as Rosenhain has pointed out*) that this inter- 

 locking is an important factor in determining the 

 strength of a piece of metal. So far as the individual 

 grain goes the strength depends upon considerations that 

 we shall deal with presently when we come to consider 

 the possible type of brickbat ; but so far as the cohesion 

 of one grain with another goes I think that it is in 

 all probability due to an interlocking of the advancing 

 arms. It may make this matter more obvious to you 

 if I show an example of crystallisation in actual process. 

 We will not attempt to crystallise a metal, because 

 we could not satisfactorily show that by means of 

 the lantern, but we shall crystallise a solution of a salt. 

 I will take common ammonium chloride, and spread a 

 film of saturated solution of that salt on a glass plate, and 

 then put it into the lantern, and you will see the fairy 

 children at work. You will see the little brickbats 

 beginning to be apparent not only at one place but at 

 many places on the surface of the glass. Crystallization 

 starts not simply from one, but from many nuclei : arms 

 are shot out : the crystals spread till they meet, and finally 

 the whole surface of the plate becomes covered with a 

 solid mass. 



* Rosenhain, Jour. Iron and Steel lust., 1904, No. 1, p. 335- 



