2014 We discuss topics related to Josephson effects where the theoretical situation is complicated, and experiments still not in a satisfactory stage. These are : 1. the appearance of a new dissipative cos ~ term, where ~ is the phase difference between the two supraconductors, 2. the microwave response of a Josephson diode at photon energies near the energy gap, 3. the effect of microwaves on critical current in the so-called Dayem bridges. INTRODUCTION TO THE JOSEPHSON EFFECT. The Josephson effect is now more than ten years old [1] and this conference on « detection and emission of electromagnetic radiation with Josephson junctions » indicates well that many applications of the effect are actively studied. The basic physics of the effect are well known now, using a two fluid model, with a pair superconducting current and a quasiparticle normal current. These two currents are not interacting as in the simple two fluid model. Consider two identical superconductors separated by a thin dielectric layer of thickness 1 lying in the XY plane. Josephson description of the barrier is based on a two dimensional field qJ(x, y, t) representing the increase in the phase of the order parameter 03C8 on crossing the barrier. is the order parameter introduced by Ginzburg and Landau whose square is proportional to the number of superconducting electron pairs. The derivatives of c are related to the electromagnetic fields E and H in the junction : where V is the voltage across the dielectric and d = 2 A + 1, 03BB the magnetic field penetration depth. The phase c determines also the supercurrent flowing across the barrier and the total current is : (*) Laboratoire associe au Centre National de la Recherche Scientifique. The first term represents the supercurrent, the second is the quasiparticle current. These equations are sufficient to explain the main phenomena predicted by Josephson and later observed. Another useful presentation has been given by Aslamazov and Larkin [2]. In their model they treat the junction as a two phase region where the amplitude of the separate phases is strongly position dependent. Thus the wave function is composed of two terms : f (x) goes to unity in superconductor 1 and 0 in superconductor 2, 03C80 is the amplitude far from the junction. Defining the current in terms of a gradient operator, the current is proportional to #E fvf sin (~1 ~2) for the supercurrent and ~f03C3V for the quasiparticle current. Aslamazov and Larkin’s representation is very useful in the case of a resistively shunted junction. Recently Notarys, Yu and Mercereau [3] have generalized this treatment, in the case of a high current density junction, by allowing the phase to be also a function of position ({J = ({J(x). In the supercurrent a new cos (({Ji ({J2) term appears and : This new supercurrent term is important for a high current density junction. This summarizes the two useful models in the classical regime. We will now present cases where the Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/rphysap:01974009010100