Almost all of us are familiar with boards that either pinch a saw blade into submission or peel away from it like a Y in the road, sometimes even cracking and splitting ahead of the blade when they are ripped along its length. This is the erratic behavior of "reaction wood," as it is known in the lumber industry, wood that is so unstable, under such internal tensions, that its reaction to a change in its neighborhood is unpredictable and dramatic-and sometimes even dangerous.
Under very special circumstances, the veneer industry prizes it, but in general, timber cutters try to avoid such trees, sawyers swear at them, buyers keep guard against them, and mill hands despise them. Every carpenter or builder will sort past them for structural or load bearing timbers because of their brash nature and inferior strength.
Economic realities force us to deal with them more often than we should have to.
Reaction wood in butt logs (the lumber term for tree trunks) occurs either from trees that grow while leaning, or from trees on steep hills or mountain sides that grow out from the bank before they grow up, in a sort of J or soft L shape. Life on a steep mountainside alone is not sufficient to cause reaction wood. Neither is wind direction nor a lopsided canopy.
On the downhill compression side, the annual rings are widely spaced, while on the uphill tension side, the rings grow tightly together. A cross section of the butt log usually reveals a pith growing so close to the tension side that the rings are dramatically eccentric. In many cases, the wood in seasoned timber leaves a furry texture behind when it is planed, and extra great effort is required to sand and polish it.
There is a lengthy section in Hunting the Osage Bow concerned with the tension and the compression sides of logs on straight growing trees and how canopy, wind direction and other effects determine them. HOB details reasons for choosing bow wood for various purposes (slats or staves) according to these two sides of the tree, and offers some speculation to explain the differences between similar bows taken from the same butt log. However, it seems to me that with normal wood, the different skill levels between bowyers accounts for a much greater difference in the quality of a bow than the position of the stave upon a straight tree, all else being equal. In other words, our efforts are better spent refining skills than worrying about the origin of a stave.
Reaction wood is an altogether different phenomenon than the tension and compression differences noted and explored in HOB. What is true for reaction wood is unique to it. You can build an exceptional bow from the tension side of reaction wood-usually after having to wrestle it into submission-but you cannot build a decent one from the compression side no matter how skillful you are as a bowyer.
Steve Allely's chapter entitled "Western Indian Bows" in Vol. One of The Traditional Bowyer's Bible shows a drawing of a standing tree as it might have been carved up for bow wood: a straight stave section split from the trunk and a stave pried loose from its limbs. Steve mentions that some Oregon Indians preferred the thin-ringed wood from limbs while others recommended thin rings from the windy side of trees or from the uphill side of trees, all this to avoid compression wood, which he noted is weaker in yew than tension wood. I read right over this paragraph ten years ago because it did not dovetail with what I knew about reaction wood and the value of thin growth rings, because it didn't distinguish between reaction wood and regular tension and compression wood, and because I had access to hardwoods but not to yew.
Somewhere else in my memory is a book drawing of a stave lifted from a limb which grew in the shape of a neatly profiled cupid's bow. The caption explained it as an aboriginal method of deriving staves. I can't identify the source at the moment, but I overlooked it, too, thinking to myself that dried reaction wood could hardly resemble the original limb profile even if the wood held up under the strain of bent limbs. I didn't understand the difference between the two kinds of reaction wood, tension and compression, nor the differences in reaction wood between softwoods and hardwoods.
However, shortly after he began experimenting with vine maple, John Strunk made an observation that sent me down a different path. He told me that the stuff usually grows in a tangle, often horizontally and always under tension, and that it had to be split mindfully because only the skyward section of each piece made a bow, whereas the underneath piece of the split was worthless as bow wood. I started wondering about these implications in other bow woods, and eventually the drawing in Steve's chapter began to gain some credibility. John was cutting reaction wood, and an outspread tree limb such as the one Steve illustrated grows as reaction wood.
This diverse information started coming together and pointing somewhere for me while I was checking out information for HOB in R. Bruce Hoadley's book, Understanding Wood. There I learned that an important physical difference exists between the tension and compression sides of reaction wood, a difference that had possible consequences for bow wood.
Reaction wood from the compression side is weaker than normal because it contains less cellulose than normal and because, as Hoadley points out, the cellulose chains are not as parallel to the long direction of the cells. However, reaction wood from the tension side is stronger than normal for the opposite reasons-because the wood contains more cellulose than normal and because the greater number of cellulose chains lie even more parallel than normal.
In softwoods, the compression side of reaction wood undergoes as much as 20 times the normal amount of longitudinal shrinkage when drying, and the shrinkage is uneven, resulting in warp and internal stresses. In hardwoods, the situation exactly reverses, and uneven and abnormal shrinkage by as much as 20 times occurs from the tension side of the tree.
What I gained from Hoadley was that stronger than average bow wood should be available from the tension side of reaction wood from hardwoods such as osage, even though such staves likely would be wild and unpredictable as they cured, whereas stronger than average bow wood should come from the tension side in softwoods such as yew, which should also cure along tamer routes. I supposed even that tension reaction wood from such conifers and softwoods as Western Indians used might even retain much of its profile if a cupid shaped bow came from such tree limbs. What I really gained, however, was the understanding that though a warped and wound construction grade two by four meant that its wood was brash and worthless, a warped and wound osage stave could be evidence of the mother lode.
John Strunk sent me several vine maple staves to play with when he started fooling with the stuff. I was struck by the tremendous amount of reflex into which they had contorted, a result, I figured later, according to my reading in Hoadley, of abnormal longitudinal shrinkage caused by reaction wood in tension pulling the stave into reflex, as much as 7 to 8 inches of it in a stave 60 inches long.
There is little doubt in my mind that such staves are indeed more resilient than normal staves from upright trees and saplings, even when compression and tension sides are identifiable in the latter. Consider that the act of bracing John's vine maple staves moves the tips about 14 inches from their original position. Translated into limb strain, the effect of hunting with one of these short bows braced for a day seems roughly equivalent to carrying around an average bow at three quarters draw. Would you have any right to expect such a bow to show anything resembling the cast in the evening that it held in the morning? Or to relax anytime soon to its original, unbraced position? Or to keep from degrading quickly?
There are downsides to building bows from hardwood reaction wood. The first problem shows up in the drying process. If you are a finesse person by temperament, one who waits for a bow to magically reveal itself without coercion, you will have a long wait with reaction wood. Occasionally, one will behave, but almost always such staves will test your resolve to make bows. I've re-corrected some staves several times with steam baths, staves that just kept twisting and reflexing excessively as they dried. Sometimes I've cut a stave into billets and spliced it at the handle into a more manageable profile. I've even punished a few of the more outrageous examples by standing them in a dark shop corner and ignoring them.
Their unpredictable nature can also get you in tons of trouble aligning limb tips through the center of the handle (yes, alignment is important, regardless of what you read or hear). The reason is that you are working with unstable limbs. As you pare away wood, the remaining wood keeps moving, readjusting to the released tensions. I've had it happen that the more I tillered, the more the tips revolted, so that the finished bow ended up more miserably aligned than when it was newly begun. If this happens, remember that seasoned wood corrects with dry heat and green wood corrects with moist heat. You can still arrive at your destination.
Tension reaction wood is by its nature thin ringed. Thin growth rings can make excellent bows, even from "normal" wood, although there is some additional degree of difficulty involved in chasing out one ring for a back. Also, a thin ringed back is more susceptible to failure from the dings and bumps that accumulate in the field.
Perhaps one osage butt log in one hundred yields bow wood worth cutting. It's a much rarer osage tree that grows a limb long and clean enough for a short stave. A limb that is long, clean and thick enough to yield book-matched billets probably would be more trouble than it's worth. Such billets do not contain uniform reaction wood from side to side. They are images of each other along their length, but each one would lack homogeneity across its width because the crown of reaction tension wood, which they share in common along one side, is of a different cellulose composition than the sides of the tree limb, which they share in common along their other side. The wood along the sides of the tree limb approaches transition status between tension and compression, sort of like the neutral plane in bow limbs. I'd guess, therefore, that limb sister billets split down the middle of the crown and then joined at the handle would show an exaggerated tendency toward reflex and corkscrew that would probably redefine the concepts. A small diameter reaction butt log big enough for only two twin staves to be cut from its tension side reacts this way. I've learned to only take one stave from small leaning trees, and then only from the crown.
An advantage that I have noticed in using limbs for staves is that the pith occasionally centers in the limb rather than offsets to the tension side, as it does in butt logs under similar stress. When it does, the rings are wider than you might expect. A tip: make sure that you mark the up side of the limb before you cut it. I've been confused more than once when the clues that were obvious for identification on the logging site mysteriously disappeared after the limb came home with me.
It's a familiar route I traveled back to the tree limb in TBB, back through it to a validation of ancient lore and wisdom. I've traveled it many times before, always to learn, as I wrote in HOB, "...that there is nothing really new under the sun in the ways of considering or making wooden bows." If I wanted to make a very short, unbacked, durable and hard hitting hunting bow, could I do better than by following ancient footsteps to a tension stave of reaction wood?
In Hunting With the Bow and Arrow, Saxton Pope anticipated the innovations of the first half of the last century, and ceded them their separate territory. He wrote:
There are many problems in the ballistics of archery that are unsolved, waiting the experiments of modern science. Empirical methods have dictated the art so far. In target equipment and shooting there is a wide field for investigation. Our interests, however, are more those of the hunter, and less those of the physicist.
It would be presumptuous to think that from our insulated lives we can discover something new in the way of building wooden hunting bows and arrows, something that untold numbers of humanity gnawing on gristle for 50,000 years have not already revealed. But it is joyous fun trying something new or different, exploring this territory with eyes open and full of wonder and respect. Insights into more than bows and arrows almost always await our discovery.
Ultimately you will have to ask yourself if working tension reaction wood in hardwoods is worth the struggle for the results, and you will have to answer by what you value in a bow. I'm still experimenting with it myself, still finding out. Early results suggest that the perfect hunting bow I have been questing after these many years may lie hidden in this imperfect wood.
CAUTION: Cutting reaction wood in standing timber of any significant size is extremely dangerous. Do not cut such a tree unless you are familiar with the plunge technique used for felling veneer logs, where the pie is cut less than half way to the center (in this case, on the compression side), a uniform hinge remains, a plunge cut sloping toward the pie wades out the wood from the center and leaves a strap behind on the tension side while nylon wedges support the weight of the tree at the strap. Never walk around the strap side of the tree to drive wedges or to compare the hinge widths. Never turn your back to a cut tree while working on it, and when moving around it, always from the pie side, keep a stiff-armed hand on the trunk, as you would with livestock. When all is ready, cutting the strap fells the tree.