“I want you, Sergeant, to give me a bit of help,” he began. “First, I want the weights of the crate and the bar of iron. Can you get them for me?” “Certainly. We’ve nothing here that would weigh them, but I’ll send them to the railway station. You’ll have the weights in half an hour.” “Good man! Now there is one other thing. Can you borrow a Molesworth for me?” “A Molesworth?” “A Molesworth’s _Pocket Book of Engineering Formulæ_. You’ll get it from any engineer or architect.” “Yes, I think I can manage that. Anything else?” “No, Sergeant; that’s all except that before you send away the crate I want to measure those nail holes.” French took a pencil from his pocket and sharpened it to a long, thin, evenly rounded point. This he pushed into the nail holes, marking how far it went in. Then with a pocket rule he measured the diameter of the pencil, the length of the sharpened portion, and the distance the latter had entered. From these dimensions a simple calculation told him that the holes were all slightly under one-sixth of an inch in diameter. The sergeant was an energetic man and before the half-hour was up he had produced the required weights and the engineer’s pocket book. French, returning to the hotel, sat down with the Molesworth and a few sheets of paper, and began with some misgivings to bury himself in engineering calculations. First he added the weights of the crate, the body, and the steel bar; they came to 29 stone, or 406 pounds. Then he found that the volume of the crate was just a trifle over 15 cubic feet. This latter multiplied by the weight of a cubic foot of sea water—64 pounds—gave a total of 985 pounds as the weight of water the crate would displace if completely submerged. But if the weight of the crate was 406 pounds and the weight of the water it displaced was 985 pounds, it followed that not only would it float, but it would float with a very considerable buoyancy, represented by the difference between these two, or 579 pounds. The first part of his theory was therefore tenable. But the moment the crate was thrown into the sea, water would begin to run in through the lower holes. French wondered if he could calculate how long it would take to sink. He was himself rather out of his depth among the unfamiliar figures and formulæ given on the subject. The problem was, How long would it take 579 pounds of water to run through seven one-sixth-inch holes? This, he found, depended on the head, which he could only guess at approximately one foot. He worked for a considerable time and at last came to the conclusion that it would take slightly over an hour. But that his calculations were correct he would not like to have sworn. At all events, these results were extremely promising and gave him at least a tentative working theory. But if the crate had floated from the coast to where it was found, the question immediately arose: At what point had it been thrown in?