Using wide-field photometric data from the Sloan Digital Sky Survey ( SDSS ) we recently showed that the Galactic globular cluster Palomar 5 is in the process of being tidally disrupted .
Its tidal tails were initially detected in a 2.5 degree wide band along the celestial equator .
A new analysis of SDSS data for a larger field now reveals that the tails of Pal 5 have a much larger spatial extent and can be traced over an arc of 10 ^ { \circ } on the sky , corresponding to a projected length of 4 kpc at the distance of the cluster .
The tail that trails behind the Galactic motion of the cluster fades into the field at an angular distance of 6 \fdg 5 from the cluster center but shows a pronounced density maximum between 2 ^ { \circ } and 4 ^ { \circ } from the center .
The leading tail of length 3 \fdg 5 extends down to the border of the available field and thus presumably continues beyond it .
The projected width of these tails is small and almost constant ( FWHM \sim 
120 pc ) , which implies that they form a dynamically cold and hence long-lived structure .
The number of former cluster stars found in the tails adds up to about 1.2 times the number of stars in the cluster , i.e .
the tails are more massive than the cluster in its present state .
The radial profile of stellar surface density in the tails follows approximately a power law r ^ { \gamma } with -1.5 \leq \gamma \leq - 1.2 .

The stream of debris from Pal 5 is significantly curved , which demonstrates its acceleration by the Galactic potential .
The stream sets tight constraints on the geometry of the cluster ’ s Galactic orbit .
We conclude that the cluster is presently near the apocenter but has repeatedly undergone disk crossings in the inner part of the Galaxy leading to strong tidal shocks .
Using the spatial offset between the tails and the cluster ’ s orbit we estimate the mean drift rate of the tidal debris and thus the mean mass loss rate of the cluster .
Our results suggest that the observed debris originates mostly from mass loss within the last 2 Gyrs .
The cluster is likely to be destroyed after the next disk crossing , which will happen in about 100 Myr .
There is strong evidence against the suggestion that Pal 5 might be associated with the Sgr dwarf galaxy .