e said, "The only
course is to sail south, and circumnavigate the island. In doing so we
shall effect a landing every ten miles or so, and then go into the
interior. This will thus enable us to learn all about the land. At the
same time we must survey the island, so as to learn its extent, as well
as its general shape and outline."

"But how can we survey it without the instruments?"

"That is readily done, by observing the headlands, or some special coast
line marks, and then taking the angles from those points."

"Well, that will be interesting, at least. How shall we start?"

"Do you see that point to the south which may be five or ten miles
away?"

"Yes."

"Now, examine the compass, and turn it so that the cardinal points are
directly north and south. Now sight across the face of the compass so
that you get the exact line between this point and yonder object. What
do you make it to be?"

"Why I make it out to be S. E."

[Illustration: _Fig. 5. Measuring by Triangulation._]

"That is correct. The line 1 is south by east."

"But how can we find out how long line 1 is?" asked Harry.

"Why by triangulation," said George, quickly.

"I know that, but how can we do it on sea?"

"It can be done on sea, as well as on land, but we had better go and
make the first measurement by triangulation correctly, and do this in
our subsequent measurements, unless it should be necessary to make the
measurements at sea. The plan followed on shipboard will be found
similar to the plan followed on land."

The boat was manned and the boys with a crew of the men and John made
for the shore, and together they went inland to a point marked B (Fig.
5), and sighted across to the same object C that was noted of the ship.
This, then, gave three lines, 1, 2 and 3, forming a triangle.

"If these angles are placed on a paper the distance from A to C can be
determined on the principles of proportion," remarked John.

"How is that done?"

"We will assume that the lines 1, 2, are at right angles to each other.
This is not necessary, but it happens to be so in this case. Let us
first measure the distance along the line 2, which may be any number of
inches, or feet. Suppose we call the line one inch long. Then draw the
line 1, so that it will be sufficiently long to be sure and meet the
line 3."

"Yes; I now see how it is done," remarked George, with enthusiasm. "If
the line 3 is drawn at the angle we got, when we looked at C, from B