114 Mr. Herbert Tomlinson on 



very bright red, and was allowed afterwards to cool. When 

 a certain temperature had been reached a very beautiful phe- 

 nomenon presented itself. 



Suppose that AB and CD in fig. 3 represent the two dents. 

 F». 3. 



B D 



Slight clouds were first seen at AB and CD ; these quickly 

 deepened and extended both ways, leaving a bright space at Q. 

 Immediately afterwards the clouds rolled away, beginning 

 from Q, and the whole heated portion brightened with the 

 recalescence. This phenomenon was so striking that the 

 heating and cooling were repeated several times, with appa- 

 rently no very rapid diminution of the effects of the strain. 

 After the bar had been hammered and bent considerably in 

 this and other experiments of like nature, several attempts 

 were made after straightening it to anneal it ; these attempts 

 were only partially successful. Always before recalescence 

 faint patches of cloud could be detected here and there, showing 

 that some portions were more strained than others. These 

 clouds do not seem to be formed in those specimens of steel 

 or iron which do not show the phenomenon of recalescence; at 

 least so far as can be judged from observations made with one 

 specimen of steel in which both clouds and perceptible reglow 

 were entirely absent. 



Experiment V. — Several attempts were made in this expe- 

 riment to abolish recalescence by shaking or hammering. For 

 this purpose a piece of pianoforte-steel wire, which showed 

 recalescence well, was selected. The wire, which was 1 millim. 

 thick, was suspended vertically with a weight of 1 kilo, on the 

 end. The weight was held by one hand, whilst the other hand 

 was employed in making the wire vibrate transversely whilst 

 cooling; but no amount of vibrating could prevent the reglow 

 from showing itself. A piece of the same wire was heated to 

 a very bright red in a burner; it was then removed from the 

 burner, immediately placed on an anvil, and a very sharp blow 

 was given it by a hammer. The blow did not prevent the 

 recalescence from appearing. 



In this respect, therefore, the behaviour of iron which has 

 cooled below 1000° C. without change of state seems to differ 

 from water which has cooled below 0° C. without freezing ; 

 for the latter can be made to freeze by agitation. May not, 

 however, the sudden freezing in this last case be due simply to 



