Reproduction, flowers and seeds

A moth with grey and brown wings, resting flat. — Darwin’s hawk moth (Xanthopan morganii praedicta), photographed at Andasibe, in Madagascar; Darwin had predicted the existence of such a moth from the length of the spur of the orchid Angraecum sesquipedale (photo by Bernard Dupont, 2013). Bernard DUPONT from FRANCE, CC BY-SA 2.0, source

In 1862, Darwin announced the existence of a moth that no one had ever seen.¹

It was discovered in 1903. It was named “the predicted one”.

The story usually told about it is wrong on two points — one of which changes everything.

What Darwin saw

Darwin was studying orchids and the way insects fertilise them. He was sent flowers of a species from Madagascar: Angraecum sesquipedale.

This orchid has a spur: a long, thin tube hanging below the flower, which produces the nectar.

He measured it: about twenty-nine centimetres.

He then noted a decisive detail: only the last four centimetres contain nectar.²

The reasoning

Darwin asked the fundamental question: what is a tube of such disproportionate length for?

A logical chain of reasoning follows:

If this flower exists, it must reproduce.
If it reproduces, some insect must reach its nectar.
To reach the nectar at the bottom of a twenty-nine-centimetre tube, this insect must have a proboscis twenty-five to twenty-eight centimetres long.

So he wrote: there must be a moth in Madagascar with a proboscis of such a length.³

No one had ever seen anything like it.

The astronomical comparison

Five years later, in 1867, Alfred Russel Wallace took up the prediction and made it more precise. His method deserves to be set out in detail, because it is within anyone’s reach.⁴

He did not guess. He went to measure moths in a museum.⁵

He measured the proboscis of a South American hawk moth: just over twenty-three centimetres. That of a hawk moth from tropical Africa: about nineteen centimetres.⁶

He then did a simple subtraction. A species whose proboscis exceeded these by five to eight centimetres would reach the nectar of the orchid’s largest flowers.⁷

In other words: the moth being sought is not an impossible monster. It is barely longer than those that already exist. That is what makes the prediction reasonable — and it is what Darwin had not shown.

He went on to write that such a moth may be safely predicted, and urged naturalists visiting Madagascar to search for it with as much confidence as astronomers searched for the planet Neptune — and they will be equally successful.⁸

Astronomers had calculated where an unknown planet must be, pointed their telescope, and found it.

This is neither a hunch nor a guess, but a falsifiable prediction.

The 1903 discovery

In 1903, two naturalists described a large hawk moth from Madagascar with an enormous proboscis that can exceed twenty-five centimetres.⁹

The animal was identified twenty-one years after Darwin’s death.

They named it Xanthopan morganii praedicta (“the predicted one”), as a mere subspecies of a hawk moth already known from mainland Africa.

A recent twist: in 2021, a team reopened the case by combining anatomy and DNA barcoding, and concluded that the Madagascar moth is not a subspecies but a species in its own right. It is now called Xanthopan praedicta. The predictive name, for its part, has been kept — one hundred and eighteen years after it was given.¹⁰

Two common errors in this story

1. The name does not pay tribute to Darwin

The epithet praedicta honours the prediction made by Wallace in 1867. It was he who, by comparison with real specimens, had put a figure on the proboscis being sought and identified the precise group — the hawk moths — whereas Darwin had spoken of a moth in general terms.

2. The moth was not “verified” in 1903

In 1903, the researchers described a dead specimen in a museum. The insect’s existence was proved, but not its role as a pollinator.

For nearly a century, the scientific community assumed that this hawk moth pollinated the orchid. Experimental confirmation took more than a century:

1992: first observations of the hawk moth visiting the orchid, by the German biologist Lutz T. Wasserthal.
1993-1997: publication of direct observations, infra-red films and photographs showing the moth visiting the flower.
2004: video footage showing the moth visiting the flower and carrying off its pollinia (pollen sacs).¹¹

142 years passed between the theoretical prediction and its direct proof.¹²

Summary: the scientific method in 4 stages

Stage	Date	Actor	Scientific contribution
Formulation	1862	Charles Darwin	Deduction of the existence of a long-tongued pollinator
Refinement	1867	Alfred Russel Wallace	Measurement of real proboscises, quantification of the gap, identification of hawk moths, comparison with Neptune
Description	1903	Rothschild & Jordan	Morphological identification of the museum insect
Observation	1992–2004	Wasserthal et al.	Visual and material proof of pollen transfer

The rules of pollination and of living things

A prediction is not a guess

A scientific prediction rests on a mechanical or evolutionary constraint. It is falsifiable: if thorough exploration had shown that the orchid reproduced by self-fertilisation without any insect, the hypothesis would have collapsed.

The fallacy of the exclusive pair

The popular account often sums up the situation as “one plant = one moth”. The ecological reality is a network of interactions:

Madagascar has about 30 species of long-spurred orchids.
It also has 7 species of long-tongued hawk moths, which probably form, together with these orchids, a network of linked species.¹³

Beyond the Darwin case: the second orchid

Another Malagasy species, Solenangis impraedicta — “the unpredicted one”, described in 2024 — has a spur that can reach 33 centimetres: relative to the size of its flower, it is the longest known.¹⁴

Two hypotheses remain for experiment to decide between:

The specialised mutualism hypothesis: an unknown hawk moth with an even longer proboscis exists.
The deceptive mimicry hypothesis: the flower offers no nectar and exploits the learning of moths that visit Angraecum sesquipedale. This strategy requires the deceptive flower to remain rare in its habitat so as not to educate the insects.

The general framework: flowers, seeds and population dynamics

The relationship between the orchid and the hawk moth is one instance of the general mechanisms of reproduction in angiosperms.

From flower to seed

Pollination: transport of pollen (which carries the male gametes) to the stigma (female organ) by wind, water or an animal vector (zoophily).
Fertilisation: germination of the pollen grain and union of the gametes within the ovule.
Maturation: transformation of the ovule into a seed (containing the embryo and food reserves) and of the ovary into a fruit.
Dispersal: movement of the seed away from the parent plant — carried by wind, water, or an animal that transports it in its fur or eats it and expels it elsewhere.¹⁵

This fourth stage is the exact counterpart of the first. Pollination makes pollen travel to another flower; dispersal makes the seed travel to another patch of ground. A plant is rooted to the spot, and yet everything in its reproduction is organised so that something leaves.

A plant does not bet on an individual

This is where the case of the orchid takes on its full meaning again.

A predator–prey pair or a parent–offspring pair can be described at the level of individuals. Plant reproduction cannot: it is played out in numbers. A plant produces an enormous number of seeds, the vast majority of which will never germinate — wrong place, wrong season, eaten, dried out.

What the specialisation of Angraecum buys is therefore not the survival of a flower. It is a rate: pollen deposited in the right place more often, over thousands of attempts, for thousands of generations.

The vocabulary here calls for a warning, as elsewhere on this site. People readily say that a flower “wants” to attract an insect, or that a seed “looks for” soil. These expressions are convenient, but wrong. What happens is simpler, and less reassuring: the combinations that succeeded more often left more descendants than the others. No one was aiming at anything.

Note 1

It is, however, surprising that any insect should be able to reach the nectar : our English sphinxes have probosces as long as their bodies; but in Madagascar there must be moths with probosces capable of extension to a length of between ten and eleven inches!

Darwin, Charles, On the Various Contrivances by Which British and Foreign Orchids Are Fertilised by Insects, John Murray, 1862, public domain