July 1924 saw the deep lava recovering in the crack and sealing off the water, so as to bubble up in the bottom debris of Halemaumau and push its way upward into the crevices of the island. It reached the top of Mauna Loa in 1926 and reactuated outflow at the center of the island. This migration of vents from top to bottom and back again took twelve years of fracturing, and it relieved from lava this big piece of the Hawaiian ridge. In reaching the groundwater and steamblast phase, it accomplished something which had not happened since 1790, making a supercycle of 134 years. The decade after explosion at Halemaumau was marked by small lava gushes in the bottom of the pit, bringing the depth from 1,300 feet in 1924 to 750 feet in 1934. The layers were something less than 100 feet each, and they were fed by pahoehoe conelets at the slide-rock margin. As usual, the lava was gushing up the western wall crack along the margin of the bottom magma cylinder. There was no trace of recurrence of steam blasts. Despite the excitement of actual events, experimentation continued; and I continued working on inventions for the experiments. Two approaches to our problems concerned seismic recorders which could be put in the hands of amateurs, and range finders for improving pit surveys. I had been convinced for many years that the three-component seismograph was too elaborate to be operated by volunteer school teachers or telephone operators who have other things to do. Such a seismograph records with photographic paper the north-south, east-west, and up-down motion of the ground, on a chronograph which keeps accurate time and registers a wavy line every second, so that the recording paper has to be changed and developed every day. Moreover, these instruments are for measuring distance to earthquake origins by physics of wave motion, and they have become hopelessly mathematical. Such mathematics makes for assumptions of uniformity about a rock crust which is not uniform. Qualitative science wants to know what happens at a specific rock location and wants the motion recorded by the simplest possible mechanical device. It also wants a value in number at each location, for size and direction of the first motion. This is for an earthquake, identified as one incident, over such an island as Hawaii, where the rock units are many and different. This is especially true of long periods of time when there may be no earthquakes to record. I devised a simple shock recorder, consisting of a horizontal boom of very light wood attached to a hinged weight which swung like a door, so that the boom scratched a line on a circular card which was rotated and moved along by a common alarm clock. The result was a spiral mark on the card, such that an earthquake interposed would write a zigzag opposite a place on the clock face appropriate to the time of day. All that was necessary was to remove and date the card, wind the clock once a day, and measure the zigzag.