Having decided on Mauna Kea, our task was then to settle on a site on the mountain. Our earlier work had indicated (not surprisingly) that the site quality improved as we moved up the mountain, so the summit area was a natural choice for further study. We set up two stations, one on the cinder cone complex that defines the summit area but somewhat displaced from the true summit (which we wanted to leave undisturbed) and the other on the edge of the lava plateau at a site we called “13 North” – the number indicating the altitude in thousands of feet. These two sites were identically equipped and sampled.
We also had to decide how high above the ground the primary mirror should be located since the seeing certainly would vary with height. Since it was clearly not feasible to hoist the site survey telescopes up and down to measure such a variation, we sought a surrogate measurement. Jim Westphal believed that rapid atmospheric temperature fluctuations (microthermals, as they were called) might serve this purpose and had used sensors to measure this parameter in his site evaluations for the Carnegie telescope in Chile. Jim, with his limitless enthusiasm, became swept up in the Mauna Kea adventure and offered invaluable advice and practical help. He became one of those whose encouragement in the early days of Mauna Kea was so important and so much appreciated by our fledgling band. Jim Harwood supervised the installation of the microthermal sensors at staged heights on slim 60-foot TV masts – erected with the use of a helicopter, a at that altitude that even now chills me to think of!
I have always regretted not publishing the details of our site survey. For some months in 1967/68, I worked with a young assistant, Dennis McKnight, on the results. We drafted two papers – one on the meteorology and one on the astronomical characteristics. Alas, all that survives is an incomplete draft dealing with the meteorology – along with the Preliminary Report referenced above – which of course drew on a much smaller data sample and did not cover the summit site where the telescope was ultimately erected. One delay in completing the study with McKnight arose because our seeing measurements were so very good and needed close checking. I was quite sure that the results, coming as they did from people without any background in site evaluation, would be regarded with great skepticism in the astronomical community. Also I wanted to understand how our double-beam measurements would translate to the seeing that would be experienced at a real telescope at the same site. Even in the first survey we found, for example, that our double-beam results (from a telescope a few feet above the ground and inside a windscreen) gave results consistently around 1/3 of a second of arc. This represented an exceedingly fine site since values three times higher were regarded a good seeing at established observatories!
After some months of comparative testing between the two sites we found the measured seeing indicators at the summit to be slightly superior. This, combined with its greater altitude, led us to choose it for our new telescope