The inconsistencies in the data reported by various authors on the exp
erimental variation of the densities and propagation velocities of sur
face acoustic waves depending on the Li/Nb ratio and the variations in
the compositions of the lithium niobate crystal and the lithium nioba
te melt are discussed. The inconsistency in the densities and velociti
es of surface acoustic waves is explained by the presense of tantalum
(0.51 mol %) in the starting niobium pentoxide. The descrepancies in t
he compositions of the crystals and melts are explained by the propaga
tion of the acoustic waves in a damaged crystalline layer with the den
sity 0.13 g/cm(3) lower than the density in the crystal bulk. The diff
erence between the composition of the crystal and the melt is explaine
d qualitatively by the simultaneous presence of tantalum and the defic
it of oxygen in the initial niobium pentoxide. The consideration is ba
sed on the complete isomorphism of pentavalent niobium and tantalum io
ns incorporated into the lattice of lithium niobate.