The capacitance \( C \) of a parallel plate capacitor is given by the formula:
\[
C = \frac{\epsilon_0 A}{d}
\]
where \( \epsilon_0 \) is the permittivity of free space, \( A \) is the area of the plates, and \( d \) is the distance between the plates.
For the initial configuration, the capacitance is \( C_1 \) with air as the dielectric, so:
\[
C_1 = \frac{\epsilon_0 A}{d}
\]
When a metal sheet of thickness \( \frac{2d}{3} \) is introduced between the plates, the effective distance between the plates becomes \( d - \frac{2d}{3} = \frac{d}{3} \), and the capacitance becomes \( C_2 \).
The formula for \( C_2 \) becomes:
\[
C_2 = \frac{\epsilon_0 A}{d - t + \frac{t}{K}}
\]
where \( t = \frac{2d}{3} \) and \( K = \infty \) for metals. Substituting these values:
\[
C_2 = \frac{\epsilon_0 A}{\frac{d}{3}} = 3 \times \frac{\epsilon_0 A}{d} = 3 C_1
\]
Thus, the ratio \( \frac{C_2}{C_1} = 3 : 1 \).