A RELATIONSHIP BETWEEN CIRCUMFERENCE AND WEIGHT IN TREES AND ITS BEARING ON BRANCHING ANGLES.

Murray CD. · 1927

Read A RELATIONSHIP BETWEEN CIRCUMFERENCE AND WEIGHT IN TREES AND ITS BEARING ON BRANCHING ANGLES. by Murray CD. (1927) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc2140927.txt.

Introduction

Cecil Murray, at Harvard University, measured trees in a 1927 paper linking girth and weight. One hundred sixteen measurements were made on nine kinds of trees, from aspen and hickory to oak and beech, plus branches and leaf stems. He asked how the circumference at a cut relates to the weight of all the tree beyond that point.

A straight-line relation links the logarithm of the circumference with the logarithm of the weight beyond it, for whole trees, branches, and leaf stems alike. The largest tree measured weighed far more than the smallest cutting, yet both sit on the one rule. Murray then discusses what the rule implies for the angles at which trees branch, connecting it to his minimum work work on vessels.

Catalog record

CREATORMurray CD.
YEAR1927
COLLECTIONPAPERS
FORMATTXT
FILE SIZE7 KB
LICENSEPublic Domain
RIGHTSPUBLIC DOMAIN
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC2140927

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A RELATIONSHIP BETWEEN CIRCUMFERENCE AND WEIGHT IN TREES AND ITS BEARING ON BRANCHING ANGLES By Cecil D. Murray Published in: The Journal of General Physiology DOI: 10.1085/jgp.10.5.725 ABSTRACT Observation reveals a linear relationship between the logarithm of the circumference of a tree, branch, or leaf stem, and the logarithm of the weight of the tree, branch, or leaf. The bearing of this on the angles of branching in trees is discussed. A RELATIONSHIP BETWEEN CIRCUMFERENCE AND WEIGHT IN TREES AND ITS BEARING ON BRANCHING ANGLES. By CECIL D. MURRAY. (FromHarvard University, Cambridge.) (Accepted for publication, February 21, 1927.) When a tree, at a point where the circumference is Co, divides into two branches (cl and ct), what relationship exists between co and cl + c~? In order to answer this question, which has a definite bear- ing on problems of tree form, it is convenient to investigate first the relationship between the circumference at some point and the weight, w, of all the parts of the tree peripheral to this point. Accordingly measurements were made, 116 in all, on nine kinds of trees; namely, aspen, bitternut, hickory, oak, ash, maple, cedar, hornbeam, and beech. The largest tree measured had a circumference of 56.4 cm. where cut, and the whole tree weighed 120 kg. The smallest measurements were made on the stems of leaves,--for example, circumference of stem = 0.25 cm., weight of leaf = 0.18 gm. All the data thus obtained are included in Fig. 1. Our procedure was of the simplest character. Whole trees of varying size, or branches, or leaves, were taken entirely at random from the vicinity. The only criterion of selection was that the specimen should not appear to have been recently injured. The circumference was measured by a tape, encircling the bark, at the point of section; or, for small specimens, the diameter was measured by calipers and the circumference subsequently calculated. The specimen was then weighed on one of three balances according to size. The season was midsummer, 1926; the place, Grindstone Island, N. Y. Plotting logarithms; i.e., log (weight in gin.) vs. log (circumference in cm.), the points fall close to a straight line. A statistical treat- ment yields the following numerical characteristics: Mean value of log c in the observations =- O. 161 " " " logw" " " ~ 1.250 725 The Journal of General Physiology 726 RELATION BETWEEN CIRCUMFERENCE AND WEIGHT Standard deviation of log c = 0. 507 " " " log w -- 1. 263 Mean product of simultaneous deviations ---0.637 Correlation coefficient = 0.99 From these figures one obtains, for the best linear relation between log w and log c, the equation: log w = 2.49 log c + 0.850; or, ~v ~= 7.08 c ~.,9 (1) 1.5 /. j -g,. . *Z/'* * Y / / // .// 0 1 Z ~ 4 6 1o~e1~htm ~) FIG. 1. The line is drawn according to equation (2). The trend toward a cube law relation among the observations appearing in the lower left portion of the chart may be significant. The probable error of log w is 4- 0.08,--i.e., if from the measured circumferences estimates of the weight are calculated by equation (1), then half of the actual observed weights fall within the limits +20 per cent and -17 per cent of the calculated values. It will be seen from Fig. 1 that the error is greater for the small pieces, and less CZCIL V. mmaAY 727 for the large pieces. We observed also that deviations from equa- tion (1) occurring in various parts of the same individual tree are suffi- ciently large to mask any systematic differences between the different kinds of trees that were studied. Furthermore, the equation holds as well for stems bearing nuts as for stems bearing leaves. Returning to the opening question, the solution, inherent in equa- tion (1), is given by the relation: cos'(' ffi cl 1"(9 -I- c~1'*' (2)* This follows from the fact that, if a main stem or trunk is cut near a point of branching and weighed (w0), and then if the branches are weighed separately (wl and w~), w0 must equal wl W w2. Equation (2) describes, for the class of trees studied, one special characteristic of branching. The exponent 2.49, being greater than 2, indicates, for example, that the total cross-sectional area of the branches becomes progressively greater at each branching. To ex- press this property we may say that trees follow statistically a "2.5 power law of branching." Another characteristic of branching is the equation which describes the angles of branching. In a previous paper1 this problem was dis- cussed in reference to the arterial system in animals, and the follow- ing equations, of which two only are independent, were deduced: co4 -[- c1( -- c~4 co 4 -- cl~ -~- c2 ( co ( -- ell -- c24 cos x ffi , cosy -- , cos (x+ y) ffi (3) 2 co 2 c1' 2 coscl~ 2 cl~ct* where Co, cl, and c2 are the circumferences of the main stem and the two branches into which it divides; and where x and y are the angles made by the branches (c~and c~) with the line of direction of the stem. The angle (x ~- y) is, of course, the angle included between the two branches. These equations, for our present purpose, may be con- sidered as being deduced from the assumption that the branching system connecting three points shall, for given circumferences of the stem and branches, require the least volume of wood. *Once obtained, this relation may be roughly checked by simple measure- ment, without weighing or cutting, on large trees. 1The physiological principle of minimum work applied to the angle of branch- ing of arteries, Murray, C. D., J. Gen. Physiol., 1925-26, ix, 835. 728 RELATION BETWEEN CIRCUMFERENCE AND WEIGHT If now equations (2) and (3) are combined one can, at the expense of loss of generality, solve directly for the angle as a function of some convenient ratio such as cl/co or cl/c2, The steps are shown in the previous paper. But in that paper, instead of a 2.5 power law (equa- tion (2)), a cube law (theoretically deduced for the arterial system) was used. In either case certain qualitative rules hold which describe in words the variations in the angles accompanying variations in the ratio .cl/c~, etc. An interesting and convenient illustration of these J f / J f f 0g \ 3 5 EXPONENT FIG. 2. The curve shows the relation between the angle (x + y), given that cl -- c2,calculated by equations (2) and (3), and the exponent, consideredvariable, occurring in equation (2). rules may be seen in the branching of the veins of leaves. There re- mains only to be observed the fact that, in changing from a cube law to a 2.5 power law, the calculated angles, for any given ratio of cir- cumferences, become smaller,--a fact corresponding to a difference between the branching of arteries and of trees. For example, solv- ing equation (3) for the angle (x + y) when cl/c2 = 1, we find for a cube law angle (x + y)

Read the complete text