THE PHENOMENA OF RUPTURE AND FLOW IN SOLIDS. 
197 
The foregoing conclusions are of especial interest in their relation to the theory of 
Rosenhain,* on which many of the properties of metals, and particularly “ season 
cracking ” under prolonged stress, are explained by supposing that the crystals are 
cemented together by very thin layers of amorphous material having the properties 
of an extremely viscous undercooled liquid. The experiments described above show 
that fluidity is not a property which can be ascribed a 'priori to such films. Hence 
if the view of Rosenhain and Archbutt were to be definitely established, it would 
be necessary to regard it, not as a theory of season cracking in terms of the known 
properties of materials, but as a deduction of the properties of the intercrystalline 
layers from the phenomena of season cracking. Looked at in this way, it would be 
of extreme interest, for it would show that the molecular arrangement of the inter¬ 
crystalline layers could not be of the coarse-grained type characteristic of the normal 
states of solids and liquids. 
It is clear that the foregoing theory of liquids is not free from objection, and that in 
some respects it appears to be less satisfactory than existing theories. The most 
obvious objection is that it seems to be incompatible with accepted determination of 
the molecular weight of liquids. Since, however, these experiments are based ultimately 
on kinetic considerations, the author believes that this difficulty will not in fact arise 
unless the requisite bonds between the molecules of each group are found to be sufficiently 
strong to cause appreciable modification of the average molecular kinetic energy. 
12. Summary of Conclusions. 
(1) The ordinary hypothesis of rupture cannot be employed to predict the safe range 
of alternating stress which can be applied to metal having a scratched surface. The 
safe range of an unscratched test piece appears to be slightly less than the yield range, 
but if the surface is scratched the safe range may be several times the range which 
causes yield in the corners of the scratches. 
(2) The “ theorem of minimum potential energy ” may be extended so as to be capable 
of predicting the breaking loads of elastic solids, if account is taken of the increase of 
surface energy which occurs during the formation of cracks. 
(3) The breaking load of a thin plate of glass having in it a sufficiently long straight 
crack normal to the applied stress, is inversely proportional to the square root of the 
length of the crack. The maximum tensile stress in the corners of the crack is more 
than ten times as great as the tensile strength of the material, as measured in an ordinary 
test. 
(4) The foregoing observation is in agreement with the known fact that the observed 
strength of materials is less than one-tenth of the strength deduced indirectly from 
physical data, on the assumption that the materials are isotropic. The observed 
* W. Rosenhain and D. Ewen, “ Intercrystalline Cohesion in Metals,” ‘ J. Inst. Metals,’ vol. 8 (1912) ; 
and W. Rosenhain and S. L. Archbutt, “ On the Intercrystalline Fracture of Metals under Prolonged 
Application of Stress (Preliminary Paper),” ‘ Roy. Soc. Proc.,’ A, vol. 96 (1919). 
