The sodium emissions have been observed in several new and long-period come
ts, but only for comet Mrkos 1957d (Nguyen-Huu-Doan, 1960) was a sodium tai
l detected on a Schmidt plate obtained with a objective prism. Comet Hale-B
opp 1995 O1 offered the first great opportunity to get an image of a long s
odium tail. Tt was more than 3 x 10(7) km long, defined as a third type of
tail, as it was composed only of neutral atoms (Cremonese, 1997a). After th
e discovery of the sodium tail another team announced it had observed it (W
ilson et nl., 1998), but it was soon realized they had seen a different sod
ium tail. The image of Wilson er nl. (1998) showed a very diffuse sodium ta
il superimposed on the dust tail, most likely due to the release of sodium
atoms from dust particles. It was different from the narrow tail found in t
he image obtained by the European Hale-Bopp Team and its position angle was
15-20 degrees lower. Spectroscopic observations have been performed on the
dust rail, at different beta values, and along the narrow sodium tail show
ing that the sodium emissions had very different line profiles. The analysi
s of these profiles will yield important insights into the sources in the i
nner coma and in the dust tail. This work will report on preliminary analys
is of both sodium tails and emphasize the high-resolution spectroscopy perf
ormed on the dust tail.