476 M. Fromme on the Changes produced in the Molecular 



The first point which strikes the eye upon examining this 

 table is the decrease, often considerable, which is produced by 

 repeated heating, in the density of all the kinds of iron, with 

 the single exception of electrolytic iron. 



The exception, presented by pure iron only*, suggests the 

 cause of the phenomenon ; it evidently is to be sought in the 

 varying proportion of carbon contained in the material which 

 will suffer change upon ignition. The total percentage of 

 carbon may vary, as well as the proportion between combined 

 carbon and graphite. It may be noted that none of the pieces 

 were heated longer than necessary. The smallest value of the 

 density is then often not reached after repeatedly heating. 

 This continuous decrease in density made it necessary to 

 combine the determinations suitably, before the question pro- 

 posed as to the difference in density caused by rapid cooling 

 and slow cooling could be answered. I have combined each 

 measurement with that preceding and that following it, and 

 thus obtained the numbers contained in the fourth column of 

 Table I. The intermediate members of the series 8 t and S r 

 have, however, more share in the formation of these differences 

 than the terminal numbers. Nevertheless I have taken the 

 simple arithmetical mean of all the differences to three places 

 of decimals, since it does not depend upon a knowledge of the 

 exact values of 8 r —8i. 



If we now compare with each other the mean values con- 

 tained in the last column, we find an increase in specific gravity 

 upon rapid cooling only with the iron wires a and c, the elec- 

 trolytic iron #, and the wrought iron I. We have, however, 

 only one measurement of the electrolytic iron g, and the dif- 

 ference resulting from it lies within the limit of experimental 

 error. For the wrought iron I we have two differential values; 

 but they are of small magnitude and of different sign. Hence 

 we may conclude that with the electrolytic iron (g) and with 

 the wrought iron (I) the density is sensibly independent of the 

 velocity with which the heated substance cools. 



On the other hand, an increase in specific gravity after 

 tempering is distinctly observed with the iron wires (a) and 

 (c), and with the wires (e) it is even considerable. 



All the other substances examined — that is to say, the iron 

 wires, the steel wires (d) and (e), the steel bar (/), the elec- 

 trolytic iron (]i), the cast iron (grey pig-iron) (i) and (k), and, 

 lastly, the white pig-iron (m) — all suffered a decrease in specific 



* With the piece 7i v on the contrary, the specific gravity increased upon 

 repeated heating. To begin with, it was decidedly smaller than that of 

 the other two pieces, hut upon repeated heating became continually more 

 nearly the same. 



