If $\sum _ { i = 1 } ^ { n } \left( x _ { i } - a \right) = n$ and $\sum _ { i = 1 } ^ { n } \left( x _ { i } - a \right) ^ { 2 } = n a , ( n , a > 1 )$, then the standard deviation of $n$ observations $x _ { 1 } , x _ { 2 } , \ldots , x _ { n }$ is\\
(1) $a - 1$\\
(2) $n \sqrt { ( a - 1 ) }$\\
(3) $\sqrt { n ( a - 1 ) }$\\
(4) $\sqrt { ( a - 1 ) }$