C. Barus and V. Strotihal — Strain-effect, etc. 439 



Art. XLII. — On the Strain-effect of Sudden Cooling exhibited by 

 Glass and by Steel ;* by 0. BARUS and V. Strouhal. 



Introductory. — In the foregoing number of the Journal we com- 

 municated a series of results on the structure of steel of a given, 

 kind, tempered hard. They showed, in general, that from the 

 circumference to the axis of a quenched non-fileable steel cylin- 

 der, hardness continually diminishes ; that the first fileable strata 

 are encountered at a distance of about l cm below the surface. 

 Moreover, as hardness decreases, the density of the elementary 

 conaxial cylindrical shells increases and in proportion as the 

 layers become more and nearly or easily fileable the density is 

 found to approach the density of soft steel. From an examina- 

 tion of rods of different diameters (2 om to 3 cm ) it appears, steel 

 of a given kind presupposed, that the hardness at a point is 

 essentially dependent on the distance of the point below the 

 surface. The rate at which sudden cooling takes place must 

 be similarly conditioned. Hence it is permissible to associate 

 the one phenomenon with the other and to state that the hard- 

 ness in a given point below the surface is dependent on the rate 

 at which cooling there takes place. This point of view is sug- 

 gestive : structure investigations may be made to furnish us 

 the best means we know for the comparison of hardness and 

 rate of cooling. 



The results cited apparently make sad havoc with physical 

 theories of temper. They seem indeed to be fatally at variance 

 with the usually accepted views, viz : that in hard tempered 

 steel an abnormally dense shell is arched around an abnormally 

 rare core, the two states of strain mutually conditioning each 

 other. The data may even be looked upon as furnishing evi- 

 dence sufficient to prove the total absence of strain. It is the 

 object of the present paper to investigate in what measure this 

 evidence is critically sufficient; if it be insufficient, to state the 

 nature and relations of the strain-effect of quenching somewhat 

 more clearly than we were able to do in our earlier work. 



In our little book on the iron-carburetsf we endeavored to 

 contrast the respective merits of chemical and physical theories 

 of temper, using for this purpose all the data known to us, as 

 well as special results of our own. This deduction seemed 

 warranted, that hardness and strain are distinct and separate 

 properties of a tempered iron-carburet ;% that they need not 

 necessarily coexist. Leaving therefore hardness pure to be ex- 



* Communicated with the permission of the Director of the Geological Survey. 

 fBull. U. S. Geolog. Survey, No. 14, 1885, p. 76. \ Ibid., p. 103, 197. 



