478 M. Fromme on the Changes produced in the Molecular 



with which Einmann experimented, contains only a small 

 quantity of carbon; and the designation of the steel examined 

 by H. Meyer as "soft" also points to a small percentage of 

 carbon. 



Whether the view which I have expressed, chiefly by way 

 of surmise, is correct or not must be decided by more extended 

 experiments, accompanied by analysis of the kinds of iron 

 examined. 



Mousson, in his Lehrbuch der Physik, has endeavoured to 

 explain the change in density produced by tempering as purely 

 mechanical, assuming that the initial cooling of the exterior 

 prevents the interior portions from contracting, as they other- 

 wise would do upon the fall of temperature. Hence the 

 exterior layers are left in a condition of forcible compression, 

 and the interior in a condition of forcible extension. Hence, 

 according to Mousson, three things would follow : — first, 

 an increased density and hardness of the superficial layers ; 

 secondly, a diminished density and hardness of the interior ; 

 and, thirdly, an increase in volume. 



Let us for the present disregard the first two, and concern 

 ourselves only with the third result. 



This conclusion is in opposition to the results obtained with 

 the iron wires (a) and (c), as well as the observations of Ein- 

 mann and Meyer. Any such purely mechanical explanation 

 of the processes involved in tempering can certainly not be 

 maintained. But, since Mousson's explanation takes no 

 account of chemical changes during the process of tempering, 

 and since it can be shown that such changes occur in iron 

 containing carbon, it seemed possible that the above explana- 

 tion might at least be applicable to pure iron ; and in fact 

 pure iron does exhibit on the whole an increase of volume 

 upon tempering. But, as already pointed out, the mean 

 values obtained cannot be considered decisive upon this point. 

 I have further examined platinum, but did not succeed in 

 showing any change in volume. 



On this account I am of opinion that a change of volume, 

 as a result of rapid cooling, only takes place with good con- 

 ductors, such as the metals, if chemical changes occur at the 

 same time. On the other hand, I do not at all deny the pos- 

 sibility of such changes with substances of small conducting- 

 power for heat, such as"glass. 



2. On the Distribution of Density and Hardness in Tempered 



Steel. 

 I have made experiments upon the distribution of density 

 and hardness only with one kind of steel, that marked /. 



