So I understand that the bessel functions of the first kind are the ones that satisfy this equation:

$$x^2\frac{d^2y}{dx^2}+x\frac{dy}{dx}+(x^2-\alpha^2)y = 0$$

and the result is a linear combination of the bessel functions of the first and second kind.

equation(1): $$ A J_a(x) + B Y_a(x) $$

Now let: $x = iv$

$$ \dfrac{dy}{dx} = \dfrac{dy}{dv} \dfrac{1}{i} \\ \dfrac{d^2y}{dx^2} = -\dfrac{d^2y}{dx^2}$$

Substituting in the original equation we get:

$$(iv)^2(-1)\frac{d^2y}{dv^2}+(iv)\dfrac{1}{i}\frac{dy}{dv}+((iv)^2-\alpha^2)y = 0 \\ v^2\frac{d^2y}{dv^2}+v\frac{dy}{dv}- (v^2+\alpha^2)y = 0 $$

This is the equation which has solutions the modified bessel functions.

Is equation 1 with x = iv a solution to this equation? ( I think it is but not sure )

and the second equation is: Why is this then not true?

$$ J_a(ix) = I_a(x) \\ Y_a(ix) = K_a(x) $$


1 Answer 1


Your computations are correct. Recall, however, that the space of solutions is two-dimensional: there are infinitely many ways to choose a basis there. We choose the basis as $J_\alpha$ and $Y_\alpha$, which are real-valued functions on the real axis.

Similarly, for the modified Bessel's equation we choose the basis $I_\alpha, K_\alpha$ of real-valued functions. These are linear combinations of $J_\alpha(ix)$ and $Y_\alpha(ix)$, which themselves are not real-valued.

Specifically, $$I_\alpha(x) = e^{-\alpha \pi i/2}J_\alpha(ix)$$ and $$K_\alpha(x) = \frac{\pi}{2}e^{(\alpha+1) \pi i/2}(J_\alpha(ix)+i Y_\alpha(ix))$$


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.