It began in the swimming pool at Glen Ellen. Between swims it was our wont to come out and lie in the sand and let our skins breathe the warm air and soak in the sunshine. Roscoe was a yachtsman. I had followed the sea a bit. It was inevitable that we should talk about boats. We talked about small boats, and the seaworthiness of small boats. We instanced Captain Slocum and his three years’ voyage around the world in the Spray. “The crews of vessels remaining any considerable time in the Solomons find wounds and sores liable to change into malignant ulcers.”!

“Here goes another one!” A limited vocabulary means that each word shall be overworked. Thus, fella, in bêche de mer, means all that piecee does and quite a bit more, and is used continually in every possible connection. Another overworked word is belong. Nothing stands alone. Everything is related. The thing desired is indicated by its relationship with other things. A primitive vocabulary means primitive expression, thus, the continuance of rain is expressed as rain he stop. Sun he come up cannot possibly be misunderstood, while the phrase-structure itself can be used without mental exertion in ten thousand different ways, as, for instance, a native who desires to tell you that there are fish in the water and who says fish he stop. It was while trading on Ysabel island that I learned the excellence of this usage. I wanted two or three pairs of the large clam-shells (measuring three feet across), but I did not want the meat inside. Also, I wanted the meat of some of the smaller clams to make a chowder. My instruction to the natives finally ripened into the following “You fella bring me fella big fella clam—kai-kai he no stop, he walk about. You fella bring me fella small fella clam—kai-kai he stop.”

Every mail to me was burdened with the letters of applicants who were suffocating in the “man-stifled towns,” and it soon dawned upon me that a twentieth century Ulysses required a corps of stenographers to clear his correspondence before setting sail. No, adventure is certainly not dead—not while one receives letters that begin: The Snark sailed from Fiji on Saturday, June 6, and the next day, Sunday, on the wide ocean, out of sight of land, I proceeded to endeavour to find out my position by a chronometer sight for longitude and by a meridian observation for latitude. The chronometer sight was taken in the morning when the sun was some 21° above the horizon. I looked in the Nautical Almanac and found that on that very day, June 7, the sun was behind time 1 minute and 26 seconds, and that it was catching up at a rate of 14.67 seconds per hour. The chronometer said that at the precise moment of taking the sun’s altitude it was twenty-five minutes after eight o’clock at Greenwich. From this date it would seem a schoolboy’s task to correct the Equation of Time. Unfortunately, I was not a schoolboy. Obviously, at the middle of the day, at Greenwich, the sun was 1 minute and 26 seconds behind time. Equally obviously, if it were eleven o’clock in the morning, the sun would be 1 minute and 26 seconds behind time plus 14.67 seconds. If it were ten o’clock in the morning, twice 14.67 seconds would have to be added. And if it were 8: 25 in the morning, then 3½ times 14.67 seconds would have to be added. Quite clearly, then, if, instead of being 8:25 A.M., it were 8:25 P.M., then 8½ times 14.67 seconds would have to be, not added, but subtracted; for, if, at noon, the sun were 1 minute and 26 seconds behind time, and if it were catching up with where it ought to be at the rate of 14.67 seconds per hour, then at 8.25 P.M. it would be much nearer where it ought to be than it had been at noon. But other troubles and doubts awaited me. Here is a sample of one. In south latitude, when the sun is in northern declination, chronometer sights may be taken early in the morning. I took mine at eight o’clock. Now, one of the necessary elements in working up such a sight is latitude. But one gets latitude at twelve o’clock, noon, by a meridian observation. It is clear that in order to work up my eight o’clock chronometer sight I must have my eight o’clock latitude. Of course, if the Snark were sailing due west at six knots per hour, for the intervening four hours her latitude would not change. But if she were sailing due south, her latitude would change to the tune of twenty-four miles. In which case a simple addition or subtraction would convert the twelve o’clock latitude into eight o’clock latitude. But suppose the Snark were sailing southwest. Then the traverse tables must be consulted.