“There’s a better place over on the other side of the hill,” he said, and led the way to his favorite coasting spot. But here our attention was diverted from coasting by the curious sight of a full-grown man flying a kite. We found out afterward that he was a Professor Keeler, who had made a great scientific study of kites. Professor Keeler was very affable, and we soon got acquainted with him. His kite was way up in the air, almost out of sight, and was pulling like everything. Neither Bill nor I could hold it long. But the most remarkable part of it all to me was the fact that the kite had no tail. I had heard of tailless kites made like a box, but this one appeared to be very much like the kites I had made in my younger days, and I well knew the importance of a long tail to keep such a kite steady. We asked the professor about it, and were informed that this kite was of the Malay type, which is so designed that the cloth bellies out into pockets on each side of the central stick or backbone, and these pockets balance the kite while the backbone acts as a rudder.

Finding that we were interested in the subject he gave us full instructions for making kites from 5 to 8 feet long, and these I jotted down for future use. In a 5-foot kite he said the stick should be 3/8 inch thick and 1/2 inch wide, in a 6-foot kite 7/16 inch thick and 9/16 inch wide, in a 7-foot kite 5/8 inch thick and 3/4 inch wide, and in an 8-foot kite 3/4 inch thick and 1 inch wide. On the following summer we built a 5-footer and also an 8-footer.

For the 5-foot kite we used two sticks of hickory 3/8 of an inch wide, 1/2 an inch thick, and each 5 feet long. According to directions, one stick was laid across the other at a point two-elevenths of its length from the top. Two-elevenths of 5 feet is a little less than 11 inches, and so we fastened on the cross stick 11 inches from the upper end of the backbone.
Fig. 235. Tying on the Cleats. The sticks were not nailed together, because this would have weakened the frame just at the point where it was under the greatest strain. Instead we followed the professor’s directions and tied cleats to each stick, as shown in Fig. 235, so as to form sockets. Then the sticks were laid across each other, each stick fitting into the socket of the other, just like a mortised joint. A coat of shellac on the bottom of each cleat glued it temporarily to the stick, after which it was very tightly bound with fine cord. The stick and cleats were now thoroughly shellaced. The end of each stick was tapered off to receive a brass ferrule of the kind used on
Fig. 236. Hook on the Vertical Stick. chisel handles. They can be bought at any hardware store. At the end of the backbone we fastened hooks made of brass, bent to the form shown in Fig. 236. The cross sticks were also provided with hooks, but these were double, as shown in Fig. 237, so that a hook lay on both the front and the rear side of the frame.

The frame was covered with a kind of cloth called “percaline.” The cloth was hemmed along
Fig. 237. Double Hook. each edge over heavy picture wire, and at each corner the wire was twisted around a small solid ring of brass. The rings were now slipped over the hooks on the frame and then the cross stick was bowed back by fastening a wire to the rear hooks and drawing it taut. Professor Keeler told us to tighten this bowstring until the distance from the wire to the cross stick at the center was equal to one-tenth of the length of the stick. As our sticks were each 5 feet long we tightened the wire until the cross
Fig. 238. Connection at Corner. stick bowed out 6 inches, as in Fig. 239. The belly band of the kite was fastened at one end to the lower end of the backbone and at the upper end to a wire hook at the juncture of the two sticks. The hook was fastened to the cross stick by flattening the ends and running them under the cord used for binding on the cleats (see Fig. 240). A buttonhole was made in the cloth covering to let this hook project through. The belly band was just long enough, so that it could be stretched over to one end of the cross stick, as in Fig. 241, and at this point, that is, 30 inches
Fig. 239. Bending the Cross Stick. from the upper end of the belly band, a brass ring was made fast, to which the main kite string was tied. The kite possessed the advantage that it could be quickly taken apart and folded into a small space.

An important change was made in the belly band of the kite. The lower strand was made elastic by tying it fast to a number of heavy rubber bands, as in Fig. 242. When flying the kite, if a sudden, strong puff of wind struck it, the elastic belly band would give, tilting up the lower end of the kite so that the wind passed under; but as soon as the gust had passed the rubber bands would draw the lower end of the kite back against the wind. The elastic belly band had the effect of making the kite rise almost vertically. Sometimes it would even sail square overhead. The 8-foot kite was a very powerful one. To hold it we had to use a very strong cord, the kind used by upholsterers for tying down the springs in a chair or a sofa.

Putting the Kites to Work.

Bill tested the strength of the kite once by hooking a spring scale to the kite string. The scale was made to register weights up to 25 pounds. But our kite yanked the pointer immediately past the 25-pound mark as far as it would go. We judged from this that the kite would lift at least 40 pounds. Such a pull as this it seemed a pity to waste, but how to utilize the power was a problem until one day, when the kite was soaring up on a south wind, Dutchy suggested that we tie it to one of the canoes and go sailing up-stream. We tried the trick at once, but it didn’t work very well, because the canoe was too light. The kite would drop unless there was a heavy pull on the string. We had better success with the scow, however, which provided a sufficient drag on the kite, and with the two kites to pull us we sailed a long ways up-stream, drifting down with the current when we had gone as far as we cared to.

The Diamond Box Kite.

“Materials: Four sticks, 1/4 inch thick by 5/8 inch wide by 44 inches long, for the corner sticks. Two sticks, 1/4 inch thick by 5/8 inch wide by 15 inches long, for the short spreaders. Two sticks, 1/2 inch square by about 38 inches long, for the long spreaders. Two strips of cloth
Fig. 244. Cleat for Spreader. 81 inches long, hemmed at each edge to a width of 13 inches. Whittle out twelve cleats to the form shown in Fig. 244. At the ends of the 15-inch spreaders nail cleats on each side with long wire brads, so as to form forks, as shown in Fig. 245, in which two of the corner sticks are held. The short spreaders are fastened to the corner sticks, 7 inches
Fig. 245. Corner Stick and Spreader. from the ends, with brads driven through the cleats, making the frame (as in Fig. 246). To prevent the frame from skewing off sidewise it should be braced with wire running diagonally across from one corner stick to the other. Ordinary soft stovepipe wire will do. Care must be taken to have the spreaders meet the corner sticks squarely or at right angles. Now take one of the cloth strips and sew its ends together to form a band. The end should be lapped about an inch and fastened with the sailor stitch (see Fig. 223). The same should be done to the other cross strip,
Fig. 246. The Narrow Frame. and then each band should be marked off with pencil lines at four points, all equidistant from each other. The two bands may now be tacked to the two ends of the frame with opposite pencil lines over the edges of the corner sticks, as in Fig. 247. The two remaining corner sticks are then nailed to the bands at the two other pencil lines. These corner sticks will now be braced apart by the long spreaders,
Fig. 247. Tacking on the Cloth. which are notched to the right length to stretch the cloth taut. A cleat is nailed over each notch, as shown in Fig. 248, forming forks to hold the corner pieces. The long spreaders are now forced down until they meet the short spreaders, to which they are tied with waxed string. The long spreaders may be nailed to the corner sticks by driving brads right through the cloth into the cleats and the sticks. The belly band may be fastened to any one of the corner sticks at the spreaders, and from the points where it is tied it should measure about 45 inches in length. The point where the main string should be attached to the belly band may be best
Fig. 248. Forked End of Long Spreader. determined by experiment.”

CHAPTER XXI.
THE WATER WHEEL.

“Don’t get so excited, old chap,” he replied. “That book is all right. I’m studying up some new schemes for next year’s expedition to Willow Clump Island. Why, there are lots of things in that old book that we can make.” And he proceeded to unfold his plans, sketching out some curious designs of water wheels and pumps.

By the time school closed for the summer Bill had thoroughly digested that volume, and was ready to reconstruct many of the ancient machines.

The Water Wheel.

Our first work on reaching the island was to erect a water wheel, or “noria,” as it was called in the book, in front of the camp. It had been a great nuisance to keep our filter barrel full. Every few days we would have to form a bucket brigade, passing pails of water up the line until the barrel was filled. Now Bill proposed to do away with all this bother and let the river do the work for us.

Surveying for the Water Wheel.

We first determined the height of the upper filter barrel above the level of the river. This was done with our surveying instrument, which was set level with the top of the barrel. We sighted with the instrument to a long pole that was held upright at the edge of the water. The pole had been marked off into feet with white chalk marks, and on sighting through the sight holes we found that the hairs came in line with the eleventh chalk mark. The top of the filter was, therefore, 11 feet above the level of the river. Bill figured that it would be necessary to construct a wheel about 15 feet in diameter in order to raise the water to the proper height.

First we built the towers to support the wheel. One tower was 16 feet high and the other only 10 feet. The large tower was made something like a very tall and narrow saw-horse. Two stout poles 17 feet long were flattened at their upper ends and nailed together, with the ends projecting about a foot, as shown in Fig. 251. At the bottom these poles were spaced 8 feet apart by a cross bar, and about 9-1/2 feet from the bottom a pair of boards were nailed to opposite sides of the pole to serve as supports for the axle of the water wheel. Another pair of 17-foot poles was now similarly fastened together and then the two pairs were spaced about 12 feet apart and connected at the top and bottom with boards.

At the top two smooth boards were used and these were nailed to the inner sides of the projecting ends, which were tapered off. In this manner a V-shaped trough was formed. The boards were firmly nailed together
Fig. 253. V-shaped Trough. at their meeting edges so as to prevent them from warping apart. A diagonal brace at each corner made the wedge-shaped tower very substantial. A number of cleats nailed to one of the poles provided a ladder by which we could mount to the top of the tower. The shorter tower was a three-legged affair, made of three 12-foot poles. At first two of these were flattened and nailed together at their upper ends, and they were braced at the top and bottom. The third leg was then nailed in place and braced by cross bars connecting it with the other two poles.

We were now ready to make the wheel. From Lumberville four 1/2-inch boards, each 3 inches wide and 15 feet long, were procured; also a bar of iron 3/4 of an inch in diameter and 2 feet long. At the center of one of the boards a block of wood 4 inches long and 4 inches in diameter was nailed on for a hub. A 3/4-inch hole was now drilled through this hub and the board. Holes were also drilled into the other boards at their centers. Then they
Fig. 255. The Hub. were all strung onto the bar and spaced like spokes at equal angles apart. Bill had figured it out some way that the ends of the boards should be just about 5 feet 10-1/2 inches apart. When the boards were all arranged we nailed them together at the center, and connected the ends with narrow tie boards, as indicated in Fig. 256.

Eight large tomato cans were now procured and fastened to the spokes at the ends on the inner side, that is, the side the hub was nailed to. We couldn’t very well nail on the cans, so we punched two holes in the side of each can and then secured them to the spokes by passing bolts through these holes and the boards.

The Paddles.

Our next task was to nail the receiving trough in place on the higher tower. We set up the towers on land and mounted the wheel between them with the axle resting in the crotch of the short tower and in a deep notch cut in the cross boards of the larger one. The cans on the wheel faced the larger tower, but the hub at the center and a block nailed to the larger tower spaced the wheel far enough out so that the cans did not strike the tower as they revolved. We carefully measured the distance between the spokes and the larger tower, and then built a square trough of a
Fig 259. The Receiving Trough. size to just fit into this space. This trough was nailed across the end of the V-shaped trough on top of the tower, but a notch was cut in the side so that the water would pour from the square or receiving trough into this V-shaped one. The square trough was about 8 feet long and its sides were 12 inches high; but at the ends we had to cut them down to a height of but 6 inches, so as to permit the cans to pass without hitting them.

Setting Up the Towers.

Our filter was located nearly 20 feet from the end of the river, and in order to get a good current of water to revolve our wheel we had to place it about 15 feet from shore. This necessitated building a trough line 35 feet long. Ten feet of this line were already provided in the top of the tall tower. This tower was now set up in place with the legs firmly wedged into holes excavated in the bottom of the river. The legs on the shore side were sunk a little deeper, so as to tilt the trough slightly shoreward. The outer end of the trough was about 12 feet above the level of the water. We needed but one more tower to support the remainder of the trough line. This tower was built like the first one, but was much shorter, as it was erected on land and the level of the trough at the top had to be 5 or 6 inches lower so as to make the water flow. We connected the towers by another V-shaped trough section. This we nailed to the under side of the first trough and to the inside of the second trough. The latter was then in the same way connected by a trough section with the upper filter barrel. We now rigged up our shorter tower about a foot from the taller one, wedging in the legs so that the top came level with the slotted boards of the other tower.

Mounting the Water Wheel.

Then came the task of mounting our wheel in place. We were working in a pretty strong current and found it no easy matter. In the first place, the wheel was floated down to the towers, but there it got jammed and we couldn’t lift it up. One of the paddles was broken and a bucket wrenched off before we could disentangle the wheel from the towers, and then the wheel was carried quite a distance down-stream before we could drag it in to shore.

Our next attempt was more successful. This time we anchored the wheel so that it just cleared the towers, then fastening a couple of long guy ropes to it, we raised the wheel on edge, while a boy stood on each side holding the ropes to keep the wheel steady. The anchor rope was now slowly paid out and the wheel was rolled in between the towers. This done, the wheel was lifted up and the axle rod was pushed in, with the ends of the rod resting in slots of the boards on the tall tower and in the crotch on the shorter one. To prevent the axle rod from working endwise out of its bearings, we nailed pieces of wood across the crotch and the slots against the ends of the rod. Then we cast off the anchor rope and our wheel started work, the cans dipping up the water as they were carried around by the wheel and pouring it out of the top into the receiving trough, from which the water flowed down into the filter barrel.

Cooling the Filter Barrel.

The trough line was very leaky and a great deal of water splashed out of the buckets. But for all that, within a few moments our barrel was full and overflowing. We hadn’t figured on its filling so rapidly, but we soon found a way of utilizing the surplus water. It was led to a half-barrel in which we washed our dishes, and from there it flowed through a ditch back to the river. The water for the wash barrel was taken from the top of the upper filter barrel. But we let the lower filter barrel flow over so that it would be kept wet on the outside. Our filter was fortunately placed at a point where a good breeze struck it, and we shoveled away the earth that had been piled around it so that the wind playing on the wet barrel evaporated the moisture, making the water inside very cool.

The Canvas Bucket.

This same trick was used for cooling our drinking water whenever we went off on an expedition away from camp. We had a heavy canvas bucket, the kind used on ships. We would fill this bucket with water and then hang it up in the wind. The water seeping out of the pores of the bucket would be evaporated by the wind, and this would, in a few moments, make the water inside delightfully cool. Such buckets may be bought for $1.50 to $2.00 apiece, but ours was a home-made affair, and made somewhat differently from the store kind. The canvas used was the heaviest we could find. A piece 9 inches in diameter was cut out for the
Fig. 261. Bottom of Bucket. bottom. A ring 7 inches in diameter, made of heavy brass wire, was laid on the canvas, and the cloth was turned over it and sewed down the inside of the ring. For the sides of the bucket we cut a piece 14 inches wide and 23 inches long. The upper edge was strengthened by a piece of light rope held in place by hemming the cloth over it. The lower edge was now sewed to the bottom, just inside the wire ring and then the ends of the piece were joined, completing the sides of the bucket. The bail of the bucket was formed of a piece of rope fastened to the roped upper edge of the bucket.

We thought he was fooling at first, but when he had assured us that he was in earnest, Bill told him that we needed our own plant, but we could build him a similar and even better current wheel for any amount he thought it was worth to him. The figure settled on was six dollars (a dollar apiece) for our work, Mr. Halliday paying for the material. It was not a large sum, but it seemed
Fig. 263. Mr. Halliday’s Water Wheel. a lot to us, and considering the scarcity of money in that region it was pretty generous pay. We built Mr. Halliday’s current wheel just like our own, except that the paddles were much broader, and instead of using cans for the buckets Mr. Halliday supplied us with small dinner pails. The method of fastening on the pails is shown in Fig. 263. A stick was nailed across the end of each spoke and the bail of the pail was held by a screw eye threaded into this stick. The pails would hang straight, holding all the water without spilling a drop until the receiving trough was reached. This trough was fastened high enough to strike the bottom of the pails as they went by, tipping them over and emptying them of their contents. From the trough the water ran directly into a large cider barrel and from here was carried through a pipe to Mr. Halliday’s barn. A stopcock was here provided so that he could turn the water on or off, as he desired. The use of pails was a great improvement on tin can buckets. Fully three times as much water was poured into the receiving trough, because not a drop was spilled out on the way up.

CHAPTER XXII.
THE LOG CABIN.

The logs for the house were cut from a tract of wooded land about five miles up the river, belonging to Mr. Schreiner. To be sure we could have cut the timber from our own island, but when Reddy had said something to his father about our building a log cabin, Mr. Schreiner had warned us not to cut down any of the trees without the owner’s permission. All we could learn about the owner was that his name was Smith, and that he lived somewhere in New York city. It seemed unlikely that he would ever have anything to say about our cutting down a few trees, but rather than run any risk Mr. Schreiner advised us to make use of his woods for any timber we might need. Accordingly we started out early one morning on a logging expedition. We had no apparatus for handling any logs more than 6 or 8 inches in diameter, and Bill reckoned it out that we would have to have about fifty logs of this size for the sides of the building alone. This did not mean that fifty trees had to be chopped down, because we could usually cut two logs from a single tree. As the logs would have to overlap about a foot at each corner, we had to cut the longer ones to a length of 14 feet and the others to a length of 12 feet. Aside from these we had to have several 16-foot logs for the roof. Only the straightest logs were chosen, and while Bill and Reddy wielded the axes the rest of us hacked off the small branches with hatchets and hauled the sticks down the river. Here we tied them together to make a raft.

The Log Raft.

This was done by running a pair of ropes alternately over and under the logs at each end (see Fig. 264). About fifteen were thus fastened together, and then as an extra precaution a log was laid across each end of the raft and tied fast. As soon as we had cut enough timber for our first raft, we all ceased
Fig. 264. Tying the Logs Together. work, to take a ride down the river on the logs. Two of us, armed with poles, were to do the steering. There was one spot in the river of which we were rather apprehensive. That was a bit of shallow, swift water three miles from camp. A line of rocks jutted up from the river, forming a natural dam which was broken only at the eastern end. The water swirled madly through this opening, and veering off a huge rock which lay directly in front of the gap turned sharply westward. As we neared this dam the river became deeper and deeper, until finally we could no longer reach bottom with the poles, and could not properly steer the boat. For some time we drifted helplessly round and round in the still water above the dam. Then suddenly the current caught us and we swept like a shot for the opening. The gap was quite wide, and had we only thought to provide ourselves with oars we could have steered the raft clear of the rocks below, but we were entirely at the mercy of the current, and with a terrific crash we were hurled head on against the boulder.

Shouting to Bill to let go his hold on the rock. I swam over and caught him as he drifted down, then I helped him ashore. Leaving Bill to recuperate I rushed down the bank, shouting to the others to paddle the logs over toward shore. Then I plunged in, and pulling myself up on the nearest log, paddled shoreward as we had done on the planks when shooting the rapids. In this way one by one we corralled the logs, and after tying them together again resumed our voyage down the river. We now had no swift water to fear and were able to guide the raft successfully down to Lake Placid. But here we moored it, not venturing to take it past the mill-race until we had gotten the oars from the scow and nailed on oar locks at each side and the rear, so that we could properly row and steer the raft safely to Kite Island.

The Sail-Rigged Raft.

When we went up the river again we carried the oars with us, also the sail and mast belonging to our ice boat, as there was a good breeze blowing down-stream. Our second trip was more successful. The mast was stepped in a small but solid box nailed to the logs. In the top of this box a hole was cut for the mast to fit into and then the mast was braced with guy lines. We came down the river in fine style, steering straight for the opening in the dam, and just as we were about to shoot through Reddy and I plied the oars for all we were worth on the port (left) side so as to swing the raft around past the boulder. However, we didn’t escape entirely without accident, for the raft rode up on a submerged ledge, dipping the starboard side clear under water and nearly tipping us over. But in a moment the raft had righted itself and we had smooth sailing for the rest of the way.

Building the Log Cabin.

Our third expedition completed the number of logs we required for the log cabin. Two large 12-foot logs were chosen for the foundation logs at the front and rear of the building. The logs were flattened along the bottom so that they would have a firmer bearing on the ground, and particularly on the corners, where they rested on foundation stones. Each log was now notched about a foot from the ends. The notches were 8 inches long and about 2 inches deep. Care was taken to place those on one log squarely opposite the notches on the other. A pair of 14-foot logs were now laid across the foundation logs and rolled along them until another half-turn would have dropped them into the notches (shown in Fig. 266). Then notches were cut in the 14-foot logs to correspond, so that when the final half-turn was given one notch would fit over the other, making a mortise joint (Fig. 267). When the side logs were in position notches were cut in their upper surface to receive a pair of 12-foot logs which were rolled onto them, notched and dropped into place. Then another pair of side logs were laid on, and so the work progressed. The notches in each log were cut to a depth equal
Fig. 267. Foundation Logs Fitted Together. to one-quarter the diameter of the log; that is, if the log was 8 inches in diameter the notch was made 2 inches deep, and if 6 inches in diameter it was cut to a depth of 1-1/2 inches. When the logs were laid in place no space intervened between them, as will be clearly understood by reference to Fig. 268.

We found, after a few logs had been set in place, that our cabin was growing faster at one end than at the other. The trouble was that our logs were not of uniform diameter throughout, and we had been laying the butt ends, which were larger, all at one end of the building. So we had to take down the logs and relay
Fig. 268. A Corner of the Log Wall. them with the butt end of the front foundation log at one end and that of the rear foundation log at the other. Then the cross logs were laid on with their butt ends on the small ends of the foundation logs. The next end logs were laid with their small ends on the butt ends of the cross logs, and so on, taking care never to lay the butt end of one log across the butt end of another. In this way the walls were built up evenly to a height of 3 feet.

We had planned to make a large open fireplace in the cabin, and this necessitated cutting an opening in the rear wall. But we did not want to cut the opening until the wall was built up to its full height lest it might buckle while the remainder of the logs were being placed in position. So we merely cut a piece out of the top log to make room for a saw when we were ready to cut the complete opening. As our fireplace was to be 5 feet in width, a 5-foot piece was cut out of the center of the log. Then the ends were supported by cleats nailed on each side, as shown in Fig. 269. This done the building was continued as before, but as the walls grew we found it more and more difficult to raise the logs to position. We could not lift them directly to the top of the wall, but had to roll them up on “skids”; that is, on a pair of 14-foot logs which were laid against the top of the wall. When the walls had reached a height of about 5 feet above the foundation logs, a length 4 feet 9 inches long was cut out of the top log to allow space for sawing out the front door and window, and also a 30-inch piece was cut out for the side window. Cleats temporarily held the sawed ends of the logs, while the walls were carried on up to a height of a little over 6 feet from the foundation logs.

The Roof of the Log Cabin.

Then we started laying the roof. A 16-foot log was now notched in place at each side, with its forward end projecting about 3 feet over the front of the cabin to form a shelter in front of the building. A pair of 12-foot logs were then laid in position. The next pair of 16-foot logs were laid about 20 inches in from the sides, and after a pair of the cross logs had been set in place a third pair of logs were laid about 40 inches from the sides. Finally, a single 16-foot log was set in place at the center, to serve as the ridge beam of the roof. The roof logs were all carefully tested to see if they were sound before we laid them in place, because we did not want to run any risk of the roof falling in, particularly in the winter time, when it would be heavily covered with snow. A chalk line was drawn from the ridge beam to the lower roof beam, and the cross logs were sawed off along this line, as indicated in Fig. 271. Several slabs were now procured and laid across the roof beams to serve as rafters. These rafters projected about 18 inches beyond the side walls of the cabin, so as to support the eaves. Over the rafters we laid a roofing of slabs, starting with the bottom and lapping them, as we had done on our tree house.

We were now ready to cut out and frame the doors and window openings. The front window of the cabin was to be close beside the door, so we merely widened the door opening at the top to include the window opening as well (see Figs. 271 and 272). The door was made 2-1/2 feet wide, and was cut down to the foundation logs. The window opening was cut to a depth of 24 inches. Before sawing out the opening we wedged pieces of wood between the logs along the line we were to follow with the saw, so as to keep them in place. After the opening had been made a couple of stout boards were nailed to the sawed ends of the logs at each side, to hold them securely in place and make a suitable framing for the door. The cleats were then removed. The foundation log and the one at the top of the opening were flattened, to serve as the sill and lintel of the door. Between the door and window a short post was wedged in place. This post was flattened on opposite sides, so that the door jamb could be nailed against it on one side and the window frame on the other. The side window was next cut out and framed. After it had been framed it measured 2 feet square.

That was just what we wanted. The banks Reddy referred to turned out to be genuine red shale, and soon we had ferried several scow loads of the stuff down to Kite Island. When the shale was wet it made quite a sticky mortar. The foundations of the chimney were laid in a trench about 2 feet deep, and the side walls of the chimney were carried inside of the cabin and covered the ends of the logs at the chimney opening. The side walls extended outward a distance of 3 feet, where they were joined by the rear wall of the chimney.

The Proper Way to Build a Stone Wall.

In making our chimney we could not rely on the red shale to hold the stones as firmly as good lime mortar would, so we had to be careful that each stone, as it was laid, had a firm bearing. The stones were embedded in a thick layer of mud, and if they showed any tendency to teeter we propped them up by wedging small stones under them until they lay solid. Another thing that we were very careful about was to “break joints”; that is, to keep the joints in each layer of the stones from coinciding with those in the next layer, above or below. To make sure of this we made it a point to lay a stone over each joint in the top of the wall and then to fill in the space between the stones with smaller stones. In this way the wall was made very substantial.

When the masonry had been carried up to the top of the chimney opening, a heavy timber about 12 inches wide was laid across the walls close against the wall of the building. This was to support the fourth wall of the chimney, and so we flattened its upper surface. To prevent it from catching fire it was covered with a thick plastering of mud, and then to keep the mud from cracking and flaking off we procured
Fig. 273. How to Build a Wall. a piece of tin and tacked it over the log. The tin also extended over the top log of the opening. Then we went on with the building of the chimney walls, carrying them up about a foot above the ridge of the roof. Our chimney was completed by paving the bottom with stones, well packed in mud and nicely smoothed off to make the hearth. The hearth extended about 18 inches into the cabin, and was framed with logs, as shown in Fig. 275.