Chapter X. On The Imperfection Of The Geological Record
Charles Darwin · 1872. Read the original ↗
The absence of finely graded fossil links, and the apparent sudden appearance of whole groups of species, are explained by the extreme imperfection of the geological record. Darwin argues these gaps are therefore not fatal to descent with modification through natural selection.
The claim that the record is poorly preserved and gapped is well supported by observation (B). The inferences from it, that the missing links and ancestors existed, rest mainly on the theory and on extrapolation from breeders and a few graded series. Darwin is candid about the weakest points, especially the pre-Cambrian strata, and the agency language mostly reduces to differential replacement. The tautology risk lies in how 'improved' and 'advantage' are used.
Claims, graded
- C1ObservedB
The geological record is extremely incomplete. Fossilisation needs special conditions, formations are separated by immense unrecorded intervals, only a small part of the earth has been explored, and new discoveries keep extending the known ranges of groups.
“The explanation lies, as I believe, in the extreme imperfection of the geological record.”
Evidence offered Measured strata thicknesses, denudation and faults, the barren Flysch, coastal erosion in South America, the ranges of Chthamalus, and dated discoveries (mammals, birds, Archaeopteryx, cirripedes) that overturned earlier 'first appearance' claims.
- 1Observation///Preservation conditions, erosion, gaps between formations and revised first appearances are directly observed or documented.
- 2Hypothesis//The claim is specific about the causes of gaps, though 'extreme' is not quantified.
- 3Prediction//It implies that older fossils of 'new' groups will be found, and several were, though Darwin cites them after the fact.
- 4Experiment/There are no controlled tests, only natural comparisons such as the rising South American coast and the Malay Archipelago.
- 5Replication//Many independent geologists, including opponents of species change such as Forbes, agree that formations are separated by wide intervals.
- 6Falsifiability//Rival explanations are considered, and the Pictet editions and cirripede case show how the claim could be corrected.
What would move it up Quantitative estimates of the fraction of species and time intervals preserved, for example by comparing fossil ranges with living-species data, would remove the need to judge 'extreme' by impression.
- C2InferredC
If the theory is true, a vast number of intermediate links must once have lived. They need not lie directly between two living species but between each and a common progenitor, and their absence from any one formation is expected given the record's gaps.
“So that the number of intermediate and transitional links, between all living and extinct species, must have been inconceivably great.”
Evidence offered Pigeon breeds descended from the rock-pigeon, the Planorbis and Ammonites series, the tertiary shells ranked as varieties or species, and the point that intermediates are ranked as species unless closely linked.
- 1Observation/A few graded series (Planorbis, closely related stages) are observed, but the 'inconceivably great' number is not.
- 2Hypothesis//The claim is clear and exposes the theory to a serious objection, which Darwin concedes.
- 3Prediction/The theory predicts links, but the chapter mainly explains why few are found rather than predicting where they will be.
- 4Experiment—No test is offered, other than domestic breeds as a loose comparison.
- 5Replication/Other palaeontologists report graded forms, but the chapter gives few independent cases.
- 6Falsifiability/Both the presence of a few links and the absence of many are treated as consistent with the theory, so little would count against it.
What would move it up Dense, well-dated fossil sequences from continuously deposited strata showing graded transitions between named species and between genera, found where the preservation argument says they should be.
- C3InferredC
The apparent sudden appearance of whole groups of species, including the Cambrian faunas, reflects gaps and ignorance of the record, not real sudden origins. The groups existed earlier, probably elsewhere.
“In all cases positive palæontological evidence may be implicitly trusted; negative evidence is worthless, as experience has so often shown.”
Evidence offered Repeated pushing back of first appearances (mammals, birds, monkeys, the Chalk cirripedes, teleosts), the unexplored southern hemisphere, and a hypothetical account of ancient oceans and metamorphosed older strata. Darwin admits he can give 'no satisfactory answer' for the missing pre-Cambrian strata.
- 1Observation//The revised ranges are well documented, but the proposed pre-Cambrian ancestors had not been observed.
- 2Hypothesis//It is precise enough to fail, since older ancestors should turn up, and Darwin says the Cambrian case would count against him if unexplained.
- 3Prediction//Earlier life was predicted in 1859 and Barrande, Hicks and Logan had found more, though the main ancestors were still missing.
- 4Experiment—Nothing in the chapter tests the claim under controlled conditions.
- 5Replication/Several independent discoveries moved ranges backward, but they concern different groups from the Cambrian ones.
- 6Falsifiability/Darwin names the pre-Cambrian case as a valid objection, but the remedy of unexplored regions and vanished strata is hard to test.
What would move it up Fossil ancestors of the abruptly appearing groups, or intermediates for them, in older strata of the kind and number the preservation argument says could survive.
- C4InferredB
Measured rates of denudation and deposition, together with the thickness of the strata, show that the earth's past is vastly long. Species-level change is therefore not time-limited, though Darwin concedes that estimates in years are uncertain and that physical estimates of the earth's age are in tension with his needs.
“1,000 feet of solid rock, as it became gradually disintegrated, would thus be removed from the mean level of the whole area in the course of six million years.”
Evidence offered Ramsay's measured thicknesses (about 72,000 feet in Britain), the faults, the denuded volcanic islands, Croll's sediment-load calculations, and Thompson's cooling estimates, which Darwin admits are very wide.
- 1Observation///Strata thicknesses, faults and sediment loads are measured.
- 2Hypothesis//Uniform-rate reasoning gives numerical estimates that could be wrong, and Darwin notes Croll's may be too large.
- 3Prediction/It was not used to make a risky prediction, though it conflicts openly with Thompson.
- 4Experiment/Modern erosion is measured, but there is no controlled test of deep-past rates.
- 5Replication//Ramsay, Jukes, Geikie, Croll and Lyell reach similar conclusions independently.
- 6Falsifiability//Darwin sets his estimates against Thompson's and halves or quarters Croll's figures.
What would move it up Independent absolute dating methods that agree with the sedimentation-based estimates, and a resolution of the conflict with the physical estimates of the earth's age.
- C5ExtrapolatedD
Change that breeders achieve in decades or centuries, scaled to natural populations over geological periods, makes the production of new species and eventually whole classes plausible. Darwin allows that natural change is slower and that nature has only 'unconscious' selection to compare with.
“Few men have attended with due care to any one strain for more than half a century, so that a hundred years represents the work of two breeders in succession.”
Evidence offered Domestic sub-breeds produced in a lifetime, breeds altered over two or three centuries by unconscious selection, and the strip-of-paper illustration of a million years. Darwin says there is no means to measure the time needed to modify a species.
- 1Observation//Changes in domestic breeds and graded forms such as Planorbis are observed, but not species- or class-level origins.
- 2Hypothesis/'Slower than breeders' gives no quantity that could be wrong.
- 3Prediction—No rate or outcome is predicted in advance.
- 4Experiment//Artificial selection is a real controlled comparison, but of breeds within a species.
- 5Replication//Breeding results are widely repeated, but not as speciation or the origin of body plans.
- 6Falsifiability/Darwin concedes the natural rate is unknown, so no timescale would count against the claim.
What would move it up Measured rates of morphological change in natural lineages, including fossil series across species and genus boundaries, showing that breeder-scale changes accumulate at a rate fitting the available time.
- C6InferredC
Natural selection causes new, better-adapted varieties to supplant parent forms, and the rarer intermediates are 'beaten out'. This is why intermediates are scarce now and were at the time.
“for in all cases the new and improved forms of life tend to supplant the old and unimproved forms.”
Evidence offered A general argument that intermediate varieties exist in smaller numbers and so lose in competition, and that species depend more on other organic forms than on graded physical conditions. No data are given in this chapter.
- 1Observation/Replacement of forms is seen in some cases, but the chapter reports none of supplanting of parents by improved descendants.
- 2Hypothesis//The claim is clear, but 'improved' is not defined independently of the outcome.
- 3Prediction/It predicts scarce intermediates and local varieties, but this is largely fitted after the fact.
- 4Experiment—No competition test between parent and descendant forms is offered.
- 5Replication—The chapter cites no independent repetition.
- 6Falsifiability/Darwin allows parent and descendant to coexist, so replacement is not required in every case.
What would move it up Measured fitness differences, in the field or in laboratory competition, showing intermediates being displaced by descendants, with the advantage specified independently of survival.
Reasoning check
- PossibleClaim C5
“all living species have been connected with the parent-species of each genus, by differences not greater than we see between the natural and domestic varieties of the same species at the present day”
Darwin offers real evidence of change within species (domestic breeds, Planorbis, tertiary shells), but the step to genera, families and classes is asserted from the theory, and he defers it to the next chapter. Dated fossil series or measured rates showing such changes accumulating into higher groups would settle it.
- PossibleClaim C6
“Yet these birds hold their place victoriously in the battle for life; for they exist in infinite numbers and of many kinds.”
The penguin's intermediate limb is shown to be advantageous only by the bird's abundance, and 'improved' and 'advantage' elsewhere are not defined by any trait measured separately from replacement. Darwin offers abundance as the evidence. Independent performance measures for the flipper would settle it. Darwin does say the penguin is not a real transitional grade.
- PossibleClaim C3
“In all cases positive palæontological evidence may be implicitly trusted; negative evidence is worthless, as experience has so often shown.”
The blanket rule is overstated, since Darwin's own cirripede reasoning shows a thorough search can make absence informative. His evidence, repeated revisions of first appearances, is real but concerns other groups. Fossil ancestors found in older strata for the Cambrian phyla would settle it.
What it does well
- Darwin states the strongest objection to his theory in its own terms, and calls the pre-Cambrian case 'inexplicable' and 'a valid argument against the views here entertained'.
- The sections on preservation, denudation, subsidence and the intermittence of formations rest on measurements and field observations that others could check, including his own at the South American coast.
- He corrects a common misunderstanding: intermediates lie between each species and a common ancestor, not between two living species. He shows with pigeons why direct links should not be expected.
- He gives examples of his own prediction failing and being overturned (the Chalk cirripedes) and cites revisions made after Pictet's first edition.
- The agency language ('natural selection', 'the battle for life', 'new places in the polity of nature') is in most places shorthand for differential survival and replacement, and the argument survives its removal.
Limits of this evaluation Only this chapter was evaluated, and Darwin defers evidence of transitional fossils to the next chapter and the origin of variation to earlier ones. On the focus: 'natural selection' and 'the struggle for life' here describe replacement, so the reification test is passed, and the chapter never credits selection with producing new structures, so filter-as-source is not diagnosed. The breeder comparison is used for timescale, and Darwin himself says nature has only unconscious selection, so no false analogy is diagnosed. The ad hoc charge against the 'imperfect record' is not listed because Darwin gives independent support (observed erosion, discoveries) and its predictions were partly confirmed. Later findings cut both ways. Archaeopteryx and older mammals support him, but Eozoon is no longer accepted as a fossil, the Cambrian gap remains, and the age estimates from Thompson are superseded.