Abstract: By comparing the stress-strain curves of rock samples at high temperature and high pressure with ones of plexiglas at usual temperature and axial compression, and contrasting seismic wave with ones recorded when fracture occurs in the sample, and balancing the fixedpoint strain of pre- and post-earthquake with the variation of strain field pre- and main-fracture of rock sample, we understand that slow earthquake results from critical or sub-critical fracture, or pre-slid, and it is a strain wave with low frequency. It is an inevitable result due to material yield, weakening or plastic variation. The slow precursor is a sort of credible, short-term and impending precursors, such as deformation, strains wave with low frequency before the large earthquake Clarge fracture). Since the obvious critical-or sub-critical fracture, pre-slid or obvious yield was not observed before the fracture in all rocks, therefore, there won't be slow earthquake and the wave with low frequency before all strong earthquakes. As a result、the same short-term and impending precursor is not observed before all strong earthquakes.

Key words: Slow earthquake, Slow precursor, Critical fracture, Yield, Plastic variation