This question refers to the Nyquist sampling theorem. To perfectly reconstruct a bandlimited signal with maximum frequency \(f_{max}\), the sampling frequency \(f_s\) must be at least twice the maximum frequency:
\[ f_s \ge 2f_{max} \]
This minimum sampling frequency, \(2f_{max}\), is called the Nyquist rate.
The maximum permissible time interval between samples (sampling period), \(T_s\), is the reciprocal of the Nyquist rate:
\[ T_s \le \frac{1}{2f_{max}} \]
Given the signal has a maximum frequency \(f_{max} = 2 \text{ kHz} = 2 \times 10^3 \text{ Hz}\).
The Nyquist rate is \(2f_{max} = 2 \times (2 \times 10^3 \text{ Hz}) = 4 \times 10^3 \text{ Hz} = 4 \text{ kHz}\).
The maximum permissible distance (time interval) between two samples is:
\[ T_{s,max} = \frac{1}{2f_{max}} = \frac{1}{4 \times 10^3 \text{ Hz}} = \frac{1}{4000} \text{ sec} \]
\[ T_{s,max} = \frac{1}{4} \times 10^{-3} \text{ sec} = 0.25 \times 10^{-3} \text{ sec} \]
To express this in microseconds (\(\mu\)sec): \(10^{-3} \text{ sec} = 1000 \times 10^{-6} \text{ sec} = 1000 \mu\text{sec}\).
So, \(T_{s,max} = 0.25 \times 1000 \mu\text{sec} = 250 \mu\text{sec}\).
This matches option (a).
\[ \boxed{250 \text{ } \mu \text{sec}} \]