The strain distribution in an elastic plate bent in one dimension: ori
ginally proposed by Euler and Bemoulli in 1744, is examined. On the ba
sis of fundamental considerations in thermodynamics, it is shown that
the plane of zero strain is shifted slightly from the central plane. T
his appears to invalidate the currently-accepted interpretation of exp
erimental data on heteroepitaxial stress measured using the cantilever
bending method. From comparison with experimental data: doubt is cast
on the strain distribution of Euler and Bemoulli in its application t
o the cantilever bending method.