In March and April of 1997 when the comet Hale-Bopp passed through the perigee and the perihelion it reached the phase of its prominent brightness. First near the city of Kunming, Yunnan province, then at the county of Lijiang, Yunnan, two telescopes of the same type with 35 centimeter apertures had been used to make observations for this comet. A common camera with high speed films was mounted on the Cassegrain focuses of the two telescopes to get a FOV of 23' by 16'. During 14 observational days 95 exposures of the near-nucleus image had been obtained. An erupting region shaped as a fan and a shell structure were prominently appeared on the exposures of 9 observational days.

It was indicated that the angular dimension of the fan region was always kept within about 100 degrees, but it's direction was successively varied during all of the observations. In the second ten days of March, the angle between the symmetric center of the erupting region and the symmetric center of the cometary tail was less than 120 degrees and orientated toward the dust tail side. However, in the first ten days of April, this angle changed to more than 180 degrees, i.e. the symmetric center of the erupting region started to tilt toward the ion tail side. During this phase the azimuth angle variation of the fan-shaped erupting region, on the average, was about 2.5 degrees per day, i.e. the rotation period was about 148 days. The shells brighten by the fan-shaped erupting region shown a multi-layer structure and the amount of the layers was up to 5 at most. The distances between each shell were basically the same as about 18000 kilometers on the average and there was an evident outward radial motion with a velocity about 20 - 60 m/s.

Different kinds of models have been used to interpret the observational result mentioned above. A simple model of jet-spin joint has been taken to explain some phenomena. However, this model is difficult to explain why the shell structure could be kept steadily for such a long time. A density wave model could solve the structure problem, but it is very hard to explain the contradictory of too small dust grains and too large average density.