Step 1: Understanding the problem.
The matrix \( P \) has the property that every non-zero vector is an eigenvector. This implies that \( P \) must be a scalar multiple of the identity matrix, i.e., \( P = \lambda I \) for some scalar \( \lambda \). This is because the only way for every vector to be an eigenvector of \( P \) is for \( P \) to act uniformly on all vectors, which happens when \( P \) is a scalar multiple of the identity matrix.
Step 2: Analyzing the options.
(A) \( \lambda_1 \neq \lambda_2 \): This is not necessarily true. In fact, both eigenvalues must be equal since \( P \) is a scalar multiple of the identity matrix.
(B) \( \lambda_1 \lambda_2 = 2 \): This is true. Since \( P \) is a scalar matrix, the eigenvalues \( \lambda_1 \) and \( \lambda_2 \) are both equal to \( \lambda \), and hence \( \lambda_1 \lambda_2 = \lambda^2 \). The given equation suggests that \( \lambda^2 = 2 \), so \( \lambda_1 \lambda_2 = 2 \).
(C) \( \sqrt{2} \) is an eigenvalue of \( P \): This is not true because \( P \) only has one eigenvalue, which is \( \lambda = \sqrt{2} \).
(D) \( \sqrt{3} \) is an eigenvalue of \( P \): This is also false, as the eigenvalues are equal and must be \( \sqrt{2} \).
Step 3: Conclusion.
The correct answer is B, as the eigenvalues are equal and satisfy \( \lambda_1 \lambda_2 = 2 \).