ON THE CRYSTALLINE STRUCTURE OF METALS. 
281 
pressure slightly higher than that of the atmosphere. On leaving the apparatus 
the hydrogen was burnt as a small jet. 
The specimens were observed through the window by means of “ vertical ’ illu¬ 
mination coming through the objective itself; as we were content with moderate 
magnifications (up to 100 diameters) an objective of long focus could be used. 
In spite of the precautions taken, we did not succeed in keeping the polished 
surface untarnished after a red heat had been reached ; but in the course of our 
observations an interesting phenomenon was observed. On beginning an observation, 
when the specimen was still cold, the “ ferrite ” grains could be clearly distinguished. 
If the temperature of the specimen were then slowly raised by gradually turning on 
an alternating electric current, the first visible change was a dimming of the image, 
which resulted in its becoming completely blotted out. This we supposed to be due 
to condensation of moisture on some part of the optical system, but we could not 
locate it. On raising the temperature further, the image of the crystals reappeared 
very vividly; at this point, by shutting off the reflected light, the metal could be 
seen to be just vividly red. On heating still further, the pattern was rapidly and 
totally obliterated by the tarnishing of the surface ; the metal was now dull-red. 
The red then brightened, but suddenly dark spots appeared, and spread rapidly over 
the entire field of view. The speed at which they spread could, however, be 
regulated by suitably adjusting the heating current. The spots appeared well in the 
middle of the specimen, and the apparent darkening could only be pushed to the 
extreme edges by using a considerably higher temperature. On allowing the 
specimen to cool, no change was visible, either on passing through this range of 
temperature or at any other period; nor could the phenomenon be made to recur 
until the specimen had been cooled below redness; but, if this was done, the 
phenomenon would recur indefinitely in the same specimen. It seems probable that 
the action which causes this appearance occurs in the metal itself and not merely in 
the surface film of tarnish, as details in this film remain entirely unaffected by it, 
thus giving the observer the impression that he is looking at an action taking place 
beneath a thin and partially transparent film. On repeating these observations with 
the specimen maintained in other atmospheres than hydrogen, no such phenomenon 
was observed; and this leads us to suppose that the phenomenon is a result of 
chemical reaction between the hydrogen and the iron. From its occurrence just 
above redness it would seem to correspond to the arrest-point, about 487° C., discovered 
by Sir W. Robebts-Austen (‘Alloys Research Report,’ Inst. Meehan. Engin., 1899). 
Believing that this action of the hydrogen caused the surface of our specimens to 
tarnish on heating, we sought to overcome this difficulty by observing the surface of a 
specimen heated in vacuo. The heating was again done electrically, either by passing 
the current directly through the specimen, or else by placing the specimen in the centre 
of a coil of fine platinum wire wound on a piece of terra-cotta. In both cases the 
specimen was placed in a test-tube, the electrodes passing through a sealed cork at the 
vol. cxcv.— a. 2 o 
