When a shear-induced crystallization of polymers was formed between a pair of rotating plates, there existed two possibilities in forming this shear-induced crystal: one is that the shish-kebab formed is uniformly distributed in the bulk polymer melt, and the other is that it appears only on both sides of polymer-substrate interfaces. Since the mechanism of each scenario depends on the boundary conditions of the experiment, we mainly focus on the second scenario in this study and will discuss why and when the shish-kebab grows on the polymer-substrate interfaces. Through the ex-situ polarized optical images of flow-velocity gradient plane, velocity gradient-vorticity plane, and flow-vorticity plane combined with in situ phase contrast optical images and attenuated total reflection (ATR) Fourier transformed infrared spectrum, we are able to infer how shish-kebab is distributed in the bulk. We conjectured that wall slip might be the most possible and reasonable origin for the formation of the shish-kebab on the interface which leads to the two different scenarios mentioned above. This is confirmed by the results from the shear rate dependence experiment. Shear time dependence experiment is also made to study the long shear time effect. By carefully cleaning the quartz plate used for shear experiment and comparing with the results obtained from uncleaned quartz plate, we further testified our point of view and led to some physical understanding for the formation of the shish-kebab as well as the wall slip mechanism.