We glimpsed our first nuclear induction signal one chilly night in early January 1946. The term 'nuclear induction' was coined because we were looking for a small voltage induced in the receiver coil by the precessing nuclei. The 'we' were Professors Felix Bloch, William W. Hansen, and the graduate student, Martin E. Packard. It was chilly because the experiment was set up in the cyclotron laboratory which was in a leaky, unheated, glassed-over light well in the basement of the original sandstone quad of Leland Stanford Junior University.The glimpse followed a fruitless two hours of adjustment as we tried to optimize the equipment and parameters to match the expected conditions for resonance. The experimental conditions were set on the assumption that the thermal relaxation time, T 1 , would be very long, perhaps days, and the transverse relaxation, T 2 , would be very short, corresponding to the dipole-dipole interaction of the neighboring protons.For these reasons, the sample of ordinary water was 'presoaked', i.e. polarized for 24 h in a special outrider gap of the fringe field of the cyclotron, and the radiofrequency (rf) level in the probe adjusted to be a few gauss. The strength of the magnetic field was adjusted for resonance at a value corresponding to the rf frequency of 7.76 megacycles. The 4 or 5-in. magnet was connected in series with the cyclotron magnet, thus using the cyclotron power supply and regulator. The 60 cycle sweep was broad enough amply to cover the putative linewidth of about 5 G. The radiofrequency (rf) level in the probe was adjusted to be a few gauss and set to the resonant frequency by means of a surplus US Army frequency meter.The sample was removed by Bloch from the fringe field of the cyclotron magnet and placed in the nuclear induction probe which was centered in the gap of the small magnet ( Figure 1). Nothing was seen. As a last ditch attempt, the current through the cyclotron and small magnet was taken much higher than expected for resonance and allowed to drift downward at the slow rate determined by the time constant of the cyclotron magnet.I was watching the oscilloscope screen and saw a weak signal move across the screen, much as one would expect a radar signal to do. The lineshape of the signal was unique in that it was positive on the right-hand side of the screen and negative on the left-hand side. The pattern was repeated, Figure 1 The magnet and probe assembly of the first nuclear induction apparatus used by Bloch, Hansen, and Packard at Stanford, circa January 4, 1946. The knob turned the 'paddle' to balance out the leakage. The clamps held the laminated pole pieces thus releasing the euphoria which comes to researchers when they finally succeed. Thus in a few brief minutes we were compensated for the work and disappointments of the previous three months.It was quite clear that our earlier estimate of the thermal relaxation time was several orders of magnitude too high. During the next day or two Bloch reviewed his equations and had the explanation for what we had se...
