The limit does not exist
Take the logaritm of x_n we have (1/n)(ln1+ln2+...+lnn). It is larger than ∫1n x dx (from 1 to n) This integral equals nln n-n+1. Go back to the e-power and we'll get:
>n√(n!) > e^(nln n-n+1)
This shows that the sequence tends to infinity
Note: Since n√(n!) /n=1/e, it also implies that n√(n!) tends to infinity !