Abstract: During the May 1998 great event time, there was a strong storm. This storm commenced on about 2200 UT of Mad 1. The initial phase was from 2200 of May 1 to 0300 of May 2, with a sudden commencement. The main phase was from 0300 of May 2 to 0600 of May 4. The recovery phase lasted quite a few days. Observations show that there are Pc2-Pc5 geomagnetic pulsations after sudden commencement of initial phase, and these pulsations, perhaps are due to the large scale lessening or expanding of the magnetosphere. The Beijing Obs. 1 Hz Data of horizontal component H during this storm on 1-5 May 1998 are analysed. It is found that there are a lot of Pc2-Pc3 pulsations on the first 4 hours in the main phase. The power spectrum firstly, then use the band pass filter are calculated. There are some Pc5 in the recovery phase, although in China latitudes are low: L =1-2. There are some Pc5-Pc6 in the storm day time in the India. In Beijing, the Pc5 magnitude is about 5nT. It is at 18 hourLT. The band of frequencies of Pc2 pulsation is 0.1-0.15 Hz, the band of Pc3 is 0.02-0.1 Hz, and the band of pc5 is 2.0-4.0 mHz. The storm recovery phase begins when the IMF turns less southward, with delays of about 2 hours. As Fig.2 shows, the B_z. component of IMF is turning to southward when the main phase of the storm begins, and the southward IMFs （B_z＜ 0） decrease to the minimum on 0400 of May 4, two hours later the storm recovery phase begins. The solar wind electrical field vB_z is also calculated, where v is the SW speed, the B_z is the north-south component of IMF. Some results: vB_z change suddenly when the initial phase is beginning; the main phase of the storm begins at 0300 of May 2, vB_z is decreasing; the recovery phase begins when vB_z turns to the minimum, with delays of about 2 hours are got.