Golden Spiral

Phi is the positive solution of x squared equals x plus one, which makes it exactly (1 + the square root of 5) divided by 2, or 1.6180339887 and onward without repeating. Euclid defined the same ratio geometrically as cutting a line in extreme and mean ratio: the whole is to the greater part as the greater part is to the lesser. It has the unusual property that both its reciprocal and its square are itself shifted by one, so that one divided by phi is 0.618 and phi squared is 2.618.

Its connection to the Fibonacci sequence is exact and provable rather than approximate. Take 1, 1, 2, 3, 5, 8, 13, 21, 34, each term the sum of the two before it. The ratios of consecutive terms (3/2, 5/3, 8/5, 13/8, 21/13), fall alternately above and below phi and converge on it. The same holds for any sequence built by that rule, whatever two numbers it starts from, which is a stronger and less well known result.

The deep fact is about irrationality. Written as a continued fraction, phi is one plus one over one plus one over one, and so on forever, entirely ones. That is the slowest-converging continued fraction there is, which makes phi the number least well approximated by any fraction, the most irrational of the irrationals, in a sense that can be made precise. Everything genuinely true about phi in the natural world follows from this property, not from beauty or proportion.

The names are worth dating, since they are usually misdated. Divine proportion comes from Luca Pacioli's book of 1509, illustrated by Leonardo. Golden section, goldener Schnitt, is a nineteenth-century coinage generally credited to Martin Ohm in 1835, though isolated earlier appearances are documented. It was the psychologist Adolf Zeising who did most to spread the term, together with the aesthetic claims that have travelled with it ever since.

The one large, well-established appearance of phi in the living world is phyllotaxis, the arrangement of leaves around a stem and of seeds, scales and florets in a head. Count the visible spirals in a sunflower head, a pine cone or a pineapple and you generally get two counts that are consecutive Fibonacci numbers: 34 and 55, or 55 and 89, and so on up.

The underlying quantity is the divergence angle between successive primordia as they are laid down at the growing tip, which converges on roughly 137.5 degrees, the golden angle, a full turn divided in the ratio phi.

The reason is the irrationality result. If the divergence angle were a simple fraction of a turn, say one third, then every third element would line up radially and the arrangement would leave large empty wedges. The worse the angle can be approximated by a fraction, the longer anything takes to line up, and phi is the worst-approximable number there is. The golden angle is therefore the one that never repeats and packs most evenly.

This is not folklore. Helmut Vogel in 1979 gave a simple model, place element n at angle n times 137.5 degrees and at radius proportional to the square root of n, which reproduces sunflower packing convincingly. Stephane Douady and Yves Couder in 1992 showed the mechanism experimentally, dropping magnetised droplets into a rotating dish of silicone oil and finding that a system in which new elements appear at intervals and are repelled by existing ones settles onto the golden angle by itself, as a dynamical attractor. No plant needs to know any mathematics. The angle falls out of local repulsion and steady growth.

The honest caveats: not every plant shows Fibonacci phyllotaxis, other patterns including Lucas-number spirals occur, and whether the golden angle is also optimal for light capture rather than only for packing is a separate question still argued over in the literature.

Sources A Better Way to Construct the Sunflower HeadHelmut Vogel with Phyllotaxis as a Physical Self-Organized Growth ProcessS. Douady and Y. Couder

The nautilus. The chambered nautilus shell is a logarithmic spiral, and logarithmic spirals are genuinely common in nature, because that is the shape a structure takes when it grows by adding to one end without changing shape. It is not a golden spiral. When Clement Falbo measured shells in the collection of the California Academy of Sciences he found ratios near 1.33, not 1.618; a later study of eighty shells in the Smithsonian collection put the mean nearer 1.31. Expressed as growth per turn, the nautilus roughly triples its radius in a full revolution, where a golden spiral multiplies its radius by about 6.85. Overlay the two curves and they do not resemble each other.

The Parthenon. George Markowsky's Misconceptions about the Golden Ratio, published in 1992, showed that the golden rectangles habitually drawn over photographs of the Parthenon do not fit: parts of the building fall outside the rectangle, and because the structure is not rectangular and has no unambiguous edges, a measurer can pick reference points that yield almost any ratio wanted. Defenders reply that other reference points, taken from the base of the steps rather than the columns, do give phi. That is the problem rather than the rebuttal. A ratio that depends on which edge you choose is not evidence.

The Great Pyramid. The same objection applies with more force. There are a great many available lengths, the original casing is largely gone, and ratios close to phi, to pi and to several other constants can all be extracted from the same monument by choosing differently. No Egyptian text mentions the ratio.

The two spirals. The familiar diagram of quarter-circle arcs drawn through a chain of Fibonacci squares is not a golden spiral. It is piecewise circular, each arc struck from a different centre, and its curvature jumps at every join, where a true logarithmic spiral has a single centre and curvature that varies smoothly. It converges on the golden spiral as the squares grow, so it is a good approximation, but it is not the same curve. The two are conflated almost universally, including in most published illustrations of the golden spiral.