Poisson’s remarkable calculation – a method or a trick?

Since the integrand does not have an elementary antiderivative, I cannot be evaluated directly by the fundamental theorem of calculus. The familiar computation of the Gaussian integral is via the following remarkable trick, attributed to Poisson. One forms the square of I , interprets it as a double integral in the plane, transforms to polar coordinates and the answer magically pops out. The calculation is as follows