Why Do the Spirals in a Sunflower Head Come in Numbers Like 34 and 55?

1. The Spirals in a Sunflower Head Often Come in Neighbouring Fibonacci Numbers

Look at the middle of a sunflower head. The seeds form curving rows, and the rows run in two directions at once: some curl clockwise, some counterclockwise. Counting each set gives two numbers. Often they are something like 34 and 55[1]. (This is about the seed spirals, not the yellow petals.)

Those two numbers sit next to each other in a list that begins 1, 1, 2, 3, 5, 8, 13, 21, 34, 55 and carries on. Each number is the sum of the two before it. It is called the Fibonacci sequence, and these spiral pairs are consecutive numbers from it[1].

A list of numbers does not explain anything, though. What kind of arrangement of seeds would produce it?

2. Placing Seeds One at a Time, With a Small Turn Before Each One

A simple way to picture it: stand at the centre of a turntable and put down one seed. Turn the table by a fixed angle, put down the next seed a little farther out, turn again, and keep going. The only choice you make is the angle.

In 1979 a researcher named Helmut Vogel wrote this idea as a formula: the n-th seed goes at a distance proportional to the square root of n, and at n times a fixed angle around the centre. Plotted by computer, it gives pictures that look very much like real sunflower heads[2]. This is a model that copies the look of the flower. By itself, it does not say that a plant calculates anything.

So which angle works? It matters a great deal.

3. Why a Turn of About 137.5 Degrees Leaves No Gaps

Try a turn of exactly one third of a circle, 120 degrees. After three seeds you are back where you started, so every later seed lands on one of three straight lines, with empty wedges between them. A turn that is a simple fraction of a circle, such as one quarter, has the same problem: the seeds line up in straight rays[1].

To avoid that, you want a turn that never quite repeats. A turn that works well is the “golden angle,” about 137.5 degrees. It splits a full 360-degree circle into two arcs whose lengths are in the golden ratio, roughly 1.618 to 1. Of all numbers, the golden ratio is described as the one hardest to approximate with a simple fraction, so seeds placed with this turn are less likely to line up in rays and leave gaps[1][2].

The tuning looks sharp. One source shows computer pictures in which a turn one degree away from the golden angle already spoils the even spacing[2]. The same sources link the two ideas: neighbouring Fibonacci numbers, when divided, get closer and closer to the golden ratio, and the two spiral counts in a sunflower correspond to fractions that approximate it[1]. That is the link between the turntable and the counts in the first section.

But a sunflower has no protractor. How does it find that angle?

4. A Flower Does Not Measure the Angle; the Angle Can Emerge on Its Own

In 1992 two physicists, Stephane Douady and Yves Couder, published a different answer in the journal Physical Review Letters. They showed Fibonacci-type patterns in a physics laboratory experiment and in a computer simulation. In both, new elements appear one after another through a repeating process, and the pattern organises itself[3].

In their model, nobody has to set the angle. It can come out of a simple repeating rule, and the pattern builds itself[3]. I read only the abstract, so I do not describe the rule or the experiment in more detail.

That shows the pattern can build itself. It is a separate step to show that real sunflowers follow it, and for that someone has to count real flowers.

5. In 2012, Volunteers Grew and Counted Hundreds of Sunflowers

Alan Turing, the British mathematician, became interested in plant spirals in his later years, including how Fibonacci numbers could arise as a flower develops, and he died in 1954 before finishing that work[5]. In 2012, the centenary of his birth, the Museum of Science and Industry in Manchester, England, ran a public project called Turing’s Sunflowers. Families, schools and gardeners grew sunflowers and sent in their counts or photos[4][5].

A research team led by Jonathan Swinton and Erinma Ochu published the results in 2016. They collected data on 657 sunflowers. In the most reliable part of the data, they evaluated 768 spiral counts (a count of one set of clockwise or counterclockwise spirals). Of those, 565 were Fibonacci numbers, about three in four, and 67 more had a related Fibonacci structure, such as so-called Lucas numbers or doubled Fibonacci numbers[4].

So most of the counted spirals followed the pattern. What about the rest?

6. Some Sunflowers Did Not Fit the Pattern, and Why Is Still Unexplained

The study also reported cases that did not fit. One head, number 502, had 77 spirals in one direction and 56 in the other. The 56 is a Fibonacci number, but the clean, unambiguous count of 77 is far from one[4]. Another, number 667, had overlapping families of spirals competing with each other, which made the counting itself ambiguous[4].

The Manchester team reported that some sunflowers showed beautiful spirals with no Fibonacci numbers at all[5]. The authors wrote that models representing the messy, noisy way real plants develop “may be both necessary and testable” for a full understanding[4].

The honest summary: the neat turntable and self-organising pictures explain the common pattern well, but real growth is noisier than the models, and the exceptions are worth explaining rather than dismissing as miscounts. Nobody here is saying the Fibonacci pattern is wrong. It is incomplete.

7. Try It Yourself: Count the Spirals in a Sunflower Photo

Find a clear, close-up photo of the middle of a sunflower, from a book, a seed packet, or a printout. Use a pencil to trace one clockwise spiral from the centre to the edge, then count how many such spirals cross the edge of the disc. A different coloured pencil helps you keep track. Then do the same going counterclockwise.

Do your two numbers sit next to each other in the list 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89? If you cannot decide where one spiral ends, you have just met the same problem that made head 667 hard to count. Two people counting the same photo may disagree, which is worth noticing. For the full story, read the research paper listed in the sources.

Sources

Update history