Sir

Scientific images are at the centre of recent allegations of 'bad science' (for example, the 'Indonesian coelacanth' story1). How can we identify fraudulent alterations in scientific illustrations?

We suggest that the key is to identify changes, then examine them in the light of the scientist's intent. Take, for example, the old dispute between biologists Ernst Haeckel and Wilhelm His about embryos and evolution. Haeckel and others said that different animals pass through, or 'recapitulate', similar embryonic stages2. Indeed, fish and human embryos do look similar because they share primitive features — 'symplesiomorphies' in modern terms. Professor His3 disagreed, saying that embryos show distinctive hallmarks (synapomorphies) of their species group.

The evidence on both sides included drawings of embryos2,3. Haeckel's young embryos look similar, whereas His's look different. Things turned nasty when His and others accused Haeckel of doctoring pictures4. The defence, even today, is that Haeckel's figures are only schematic: it is acceptable for schematics to show alterations that help to explain the data.

We have found evidence of sleight of hand, surprisingly, on both sides. His's deer embryo has cloven hooves, but a standard work on deer embryology5 shows no such feature. This disparity arouses our suspicion. However, we have not identified His's source, so we cannot be sure that he changed anything. His embryos are at more advanced stages than Haeckel's, though, so are not valid counter-evidence.

We can make a persuasive case with Haeckel because we have identified some of his sources. When we compare his drawing of a young echidna embryo with the original6, we find that he removed the limbs (see Fig. 1). This cut was selective, applying only to the young stage. It was also systematic because he did it to other species in the picture. Its intent is to make the young embryos look more alike than they do in real life.

Figure 1: Deleting the synapomorphies from young embryos: Haeckel's echidna embryo.
figure 1

A–C, Taf X, figs 40s, 43s and Taf XI fig. 51s, respectively, from the original drawings6. D, Haeckel's copy in the Anthropogenie , 5th edition (ref. 2 Taf XI, figs VI-VIII). The limb buds of the young embryo (boxed) were deleted by Haeckel. Significantly, the other stages were copied accurately.

Full size image

Haeckel's other intent is to support recapitulation, as revealed in his text for the young embryos: “There is still no trace of the limbs or 'extremities' in this stage of development . . . [this] proves that the older vertebrates had no feet” (ref. 2, p. 371). The altered drawings support theories which the originals did not. Therefore, these are not legitimate schematic figures.

Haeckel and His both published technical works of great importance to biology. Their dubious embryo pictures appeared in non-technical, polemical books. Ironically, there was some truth on both sides: in fact, embryos show a mixture of symplesiomorphies and synapomorphies7.