B D, asep son the ratio of voltage to frequency to 60 parts per million and found agreement with the value of hf/2e then accepted. Later they increased their accuracy sufficiently to be able to discover errors in the previously accepted value of the fundamental constants and derive more accurate estimates,(21, 22) thus carrying out to fruition an early suggestion of Pippard (unpublished) The ac supercurrent is now used to compare voltages in different standards laboratories without the necessity for the interchange of banks of standard ells. If two laboratories irradiate specimens with radiation of the same fre- quency, constant-voltage steps appear at identical voltages. The intercom- parison of frequencies can be carried out in a straightforward manner by transmission of radio signals At the end of 1963, the evidence for the existence of the ac supercurrent vas only indirect. John Adkins and I tried to observe the effect by coupling r two junctions by a short("0.2 mm )thin-film transmission line The idea was that radiation emitted by one junction would modify the charac- teristics of the other. The experiment, planned to form the second part of the thesis chapter referred to above, was unsuccessful, for reasons which are still unclear. Later, Giaever (23)was able to observe the ac supercurrent by a sim- ilar method to the one we had considered, and then Yanson, Svistunov and Dmitrenko(24)succeeded in observing radiation emitted by the ac super current with a conventional detector Finally, I should like to describe the SLUG,(25) developed in the Royal Society Mond Laboratory by John Clarke while he was a research student John was attempting to make a high-sensitivity galvanometer using the pre- iously described magnetic interferometers with two junctions connected in parallel. One day Paul Wraight, who shared a room with John, observed that the fact that one cannot solder niobium using ordinary solder must mean that if one allows a molten blob of solder to solidify in contact with niobium there must be an intermediate layer of oxide, which might have a suitable thickness to act as a tunnelling barrier. This proved to be the case. However, in John specimens, in which a niobium wire was completely surrounded by a blob of sitive to externally applied m-gh the barrier proved to be completely insen solder, the critical cu magnetic fields. It was, however, found to be sen- sitive to the magnetic field produced by passing a current through the central wire. This fact led to the development of a galvanometer with sensitivity of 10 volts at a time constant of 1 s There have been many other developments which I ha describe here. I should like to conclude by saying how fascinating it has been for me to watch over the years the many developments in laboratories over the world, which followed from asking one simple question, e physical significance of broken symmetry in superconductors?163 the ratio of voltage to frequency to 60 parts per million and found agreement with the value of h/2e then accepted. Later they increased their accuracy sufficiently to be able to discover errors in the previously accepted values of the fundamental constants and derive more accurate estimates, (21, 22), thus carrying out to fruition an early suggestion of Pippard (unpublished). The ac supercurrent is now used to compare voltages in different standards laboratories without the necessity for the interchange of banks of standard cells. If two laboratories irradiate specimens with radiation of the same fre￾quency, constant-voltage steps appear at identical voltages. The intercom￾parison of frequencies can be carried out in a straightforward manner by transmission of radio signals. At the end of 1963, the evidence for the existence of the ac supercurrent was only indirect. John Adkins and I tried to observe the effect by coupling together two junctions by a short ( ~ 0.2 mm.) thin-film transmission line. The idea was that radiation emitted by one junction would modify the charac￾teristics of the other. The experiment, planned to form the second part of the thesis chapter referred to above, was unsuccessful, for reasons which are still unclear. Later, Giaever (23) was able to observe the ac supercurrent by a sim￾ilar method to the one we had considered, and then Yanson, Svistunov and Dmitrenko (24) succeeded in observing radiation emitted by the ac super￾current with a conventional detector. Finally, I should like to describe the SLUG, (25) developed in the Royal Society Mond Laboratory by John Clarke while he was a research student. John was attempting to make a high-sensitivity galvanometer using the pre￾viously described magnetic interferometers with two junctions connected in parallel. One day Paul Wraight, who shared a room with John, observed that the fact that one cannot solder niobium using ordinary solder must mean that if one allows a molten blob of solder to solidify in contact with niobium there must be an intermediate layer of oxide, which might have a suitable thickness to act as a tunnelling barrier. This proved to be the case. However, in John’s specimens, in which a niobium wire was completely surrounded by a blob of solder, the critical current through the barrier proved to be completely insen￾sitive to externally applied magnetic fields. It was, however, found to be sen￾sitive to the magnetic field produced by passing a current through the central wire. This fact led to the development of a galvanometer with sensitivity of 10-14 volts at a time constant of 1 s. There have been many other developments which I have not had time to describe here. I should like to conclude by saying how fascinating it has been for me to watch over the years the many developments in laboratories over the world, which followed from asking one simple question, namely what is the physical significance of broken symmetry in superconductors?