176 Nature Vol. 277 18 January 1979

salmon are more similar than lung­ fish and cows. And in preferring L. B. HALSTEAD AND COLLEAGUES cladogram a, cladists mean that they REPLY: The mutual relations of infer that lungfish and cows shared salmon (an actinopterygian), lung­ a more recent common ancestor than fish (a dipnoan) and cow (a mammal, lungfish and salmon. This inference a class which can be traced back to is drawn from the fact that lungfish rhipidistian crossopterygians by way and cows share derived characters of reptiles and amphibians) are cer­ (synapomorphies such as internal tainly worthy of serious considera­ nostrils, an epiglottis, a two- tion. Gardiner et al. draw up "three chambered auricle and so on cladograms showing three possible (K esteven Proc. Roy. Soc. Viet. 59, 'more closely related' pairs". Clado­ 93; 1951 )) not found in salmon. gram a is their preferred solution, Given their reliance on overall but it is certainly not true to claim, similarity, it is not surprising that as they do, that "c was the preferred evolutionary systematists find that solution of Parrington and Halstead". many of the groups they recognise The cladists simply omit the pos­ are grades. nor that these grades sibility that none of the three pairings 'happen to coincide with the major shared a more recent common an­ classes of the vertebrates' (Halstead cestor. than any of the others, but op. cit.). Yet these vertebrate classes rather were all derived from a pre­ were recognised and named by pre­ existing common stock, a view pre­ Janvier's agnathan cladogram was a Darwinian systematists, who also used viously commended by Miles (Palaeo­ reflection of Stensio's mistaken no­ overall similarity as their guide. zoic Fishes, 1971). tion of the myxinoid affinities of the naturally enough. Overall similarity Assuming that cow, salmon and heterostracans, which has been dis­ may he modernised or dignified hy lungfish were derived from such a cussed and firmly rejected by workers calling it 'genes in common' or common ancestral stock, we would in the field (see Bioi. J. Linn. Soc. S, 'shared genotype', yet when we do still insist that lungfish are closer to 339-349, 1973; Bioi. Rev. 48, 279- have access to comparative informa­ salmon on the grounds of neither 332 , 1973 , for a summary of this tion on genes. as in globin sequences having advanced significantly beyond issue). (Goodman et al. Nature 253, 603; the common heritage of all primary The current Hennigian orthodoxy 1975; Romero-Herrera et al. Nature aquatic jawed vertebrates. In marked brings to mind the situation at the 261, 162; 1976). the biochemists contrast. the cow is separated from beginning of the present century. present their information in the form both lungfish and salmon by several Then Haeckel's Biogenic Law and of cladograms. and use the same structural grades. The figure sum­ the attendant preoccupation with genealogical concept of relationship marises these relationships. genealogies, had a stultifying in­ as Hennigians. Darwin (Origin of At the meeting Gardiner presented fluence on the whole of biology. Species: see also Nelson Syst. Zoot. a cladogram which purported to 'When the results of a science are 23, 452; 1974) wrote 'our classifica­ demonstrate that lungfish were more produced by methodological assump­ tions will come to he , as far as they closely related to tetrapods than were tions rather than by the evidence can he so made. genealogies; and will either coelacanths or certain rhipi­ available, it is not science' as A. J. then truly give what may be called distian crossopterygians. The near Cain remarked m his review of the plan of creation'. He might be structural identity of crossopterygians Hennig's book in Nature. surprised that after 120 years, some and early tetrapods makes it impos­ palaeontologists would be occupied sible for them to be separated by the L. B. HALSTEAD with the defence of pre-Darwinian uniquely specialised lungfish. Later in E. I. WHITE concepts and classifications. the meeting T. S. Westoll demolished Department of Geology, Halstead's report contains many this particular cladogram and pointed University of Reading misunderstandings and ambiguities, a out that most of the features that few of which cannot be allowed to Gardiner claimed were absent in G . T. MAciNTYRE pass unchallenged. It is not, as coelacanths were in fact present in the Department of Biology, Halstead says of cladistics, "axiomatic material Westoll himself was studying. Queens College, New York that . . sister groups are given identical ranks"; it is however a logical consequence if a formal classi­ fication is required. Nor does cladis­ tics insist on a classification which Concerning Halstead's report of the done. Halstead charges us with reflects all of the ideas embodied in meeting itself, we were surprised to 'religious fervour', but the charge the cladogram. Cladists do not read that papers delivered by two of might be reversed, for to us his state­ assemble characters subjectively into us 'were demolished within hours'. ment that cladists 'are already a hierarchy: though we suspect that Freedom of speech goes a little too entrenched in some of the major the hierarchy of which Halstead far here, and we request that Halstead museums in the world' has an offen­ speaks is a reference to character justify his assertion by saying who did sive taint of McCarthyism and the weighting, not favoured by cladists. the demolition work, and how it was witch-hunt. Halstead also accuses cladists of be­ B. G . GARDINER C. PATTERSON littling the significance of parallelism. Department of Biology, P. L. FOREY But this is outside the scope of Quen Elizabeth College, P. H. GREENWOOD cladism since the recognition of such London WB R. S. MILES an evolutionary event presupposes P. JANVIER R. P. s. JEFFERIES that we already know the phylo­ Laboratoire de Palaento!ogie, British Museum (Natural History), genetic relationships of the organisms Universite de Paris, Cromwell Road, concerned. Paris v London SW7

0028-