THE RARER METALS AND THEIR ALLOYS. 509 



zirconium the subordinate point is very high up, while in the case of 

 the thallium it is very low down. So far as my experiments have as yet 

 been carried, this seems to be a fact which underlies the whole question 

 of the strength of metals and alloys. If the subordinate point is low, 

 the metal will be weak; if it is high in relation to the main setting 

 point, then the metal will be strong, and the conclusion of the whole 

 matter is this: The rarer metals which demand for their isolation from 

 their oxides either the use of aluminum or the electric arc, never, so 

 far as I can ascertain, produce low freezing points when they are added 

 in small quantities to those metals which are used for constructive pur- 

 poses. The difficultly fusible rarer metals are never the cause of weak- 

 ness, but always confer some property which is precious in industrial 

 use. How these rarer metals act, why the small quantities of the added 

 rare metals permeate the molecules, or, it may be the atoms, and 

 strengthen the metallic mass, we do not know; we are only gradually 

 accumulating evidence which is afforded by this very delicate physio- 

 logical method of investigation. 



As regards the actual temperatures represented by points on such 

 curves, it will be remembered that the indications afforded by the 

 recording pyrometer are only relative, and that gold is one of the most 

 suitable metals for enabling a high, fixed point to be determined. There 

 is much trustworthy evidence in favor of the adoption of 1,045° as the 

 melting point hitherto accepted for gold. The results of recent work 

 indicate, however, that this is too low, and it may prove to be as high 

 as 1,061.7, which is the melting point given by Heycock and Neville ' in 

 the latest of their admirable series of investigations, to which reference 

 was made in my Friday evening lecture of 1892. 



It may be well to point to a few instances in which the industrial use 

 of such of the rarer metals as have been available in sufficient quan- 

 tity is made evident. Modern developments in armor plate and projec- 

 tiles will occur to many of us at once. This diagram (fig. 1, PI. XXVI) 

 affords a rapid view of the progress which has been made; and in col- 

 lecting the materials for it from various sources I have been aided by Mi*. 

 Jenkins. The effect of projectiles of approximately the same weight, 

 when fired with the same velocity against 6-inch plates, enables com- 

 parative results to be studied, and illustrates the fact that the rivalry 

 between artillerists who design guns, and metallurgists who attempt 

 to produce both impenetrable armor plates and irresistible projectiles, 

 forms one of the most interesting pages in our national history. When 

 metallic armor was first applied to the sides of war vessels it was of 

 wrought iron, and proved to be of very great service by absolutely pre- 

 venting the passage of ordinary cast-iron shot into the interior of the 

 vessel, as was demonstrated through the American civil war. It was 

 found to be necessary, in order to pierce the plates, to employ harder 

 and larger projectiles than those then in use, and the chilled cast-iron 



1 Trans. Chein. Soc, Vol. LXVII. 1895, p. 160. 



