Hfi/2) with the measure {exp -n' 1 S^i 2 /2 V[q(r)]dr}d^0/p(-*e/ 2 ,^e xp^The high-temperature limit (classical limit) of the above expression gives dii B~f ) eq (q)=Z-1 e~^ while in the low-temperature limit (j3-**>), from the comparison of Eq. (20) with (10) it is obvious that the Feynman process goes over to Nelson's stochastic process for the ground state, provided r is identified with the physical time of evolution of the process. In other words the Feynman formulation of quantum statistical mechanics, upon lowering of the temperature, produces the switch from the ensemble averages of the classical regime to the time averages of the groundstate process arising in the stochastic quantization. From the general point of view, this makes precise the connection between the zero-temperature limit of quantum statistical mechanics and Nelson's formulation of the quantum theory. With respect to the question originally posed in connection with quantum critical phenomena, the comparison of Eq. (19) with Eq. (3) makes it evident that the framework of stochastic quantization allows us to bypass the current notion of dimensional crossover by introducing the dynamical PACS numbers: 06.In a crossed laser-atomic beam experiment the wavelengths of the 2s-3p transitions are measured in H and D to a precision of one part in 10 9 . Our value for the Rydberg constant is Roc =109 737.315 21(11) cm" 1 . The fine-structure splittings of the Sp states in H and D are 3249.8(8) and 3251.7(7) MHz, respectively; the isotope shifts for the 2s-3p 1 / 2 and 2s-3p 3 / 2 transitions are 124260.7(7) and 124262.6(7) MHz, respectively. Our results largely agree with previous, less precise experiments and with theory.