The critical redefinition of race as a permanent inherited characteristic, the biologization of the concept, and its formal taxonomic differentiation from the categories species and variety were given coherent expression in a series of papers published between 1775 and 1788 by Kant and were concretized from the late 1790s by Blumenbach. Though a pioneer in making the conformation of the skull the key signifier of racial diversity which demanded scientific classification, Blumenbach was also sharply alert to the tension between his ethical insistence on the established doctrine of human unity and the fissive implications of taxonomy or rigid physical distinction of human groups.^[35] His anthropological vocabulary was always more systematic than Buffon's but it remained versatile through the three Latin editions of De Generis Humani Varietate Nativa, 'On the Natural Varieties of Mankind' (1775, 1781, 1795), and only began to harden along racial lines in works published in German at the end of the century (1798, 1806).^[36]

Particularly in the four decades after 1766, scientific voyaging in Oceania supplied much of the empirical material on which comparative human anatomy and the nascent science of anthropology fed.^[37] This section and part of the next probe the considerable significance of representations and anatomical specimens of Oceanian people in the articulation of Blumenbach's racial thinking. The footnotes to his anthropological writings draw heavily on travel literature, especially accounts of the recent expeditions to the South Seas. In the index to the 1865 English translation of his major works (Bendyshe 1865), no traveller scores more references than do four men with strong Oceanic credentials: Joseph Banks (1743-1820), the chief naturalist on Cook's first Pacific voyage of 1768-71 and President of the Royal Society from 1778; John Hawkesworth (1715?-1773), editor of the first published narrative of Cook's first voyage; and the two Forsters. Blumenbach was an assiduous correspondent, friend, and collaborator of Banks over more than thirty years. He corresponded with both Forsters and was related by marriage to the son. Blumenbach personally set up and developed the fine Cook-Forster ethnographic collection still held in the Institut für Ethnologie at the University of Göttingen and regularly used it for illustrative purposes in his classes on natural history.^[38]

It was precisely the uneven advent of novel materials from 'novus orbis australis', 'the new southern world', that forced Blumenbach to expand and modify his initial quadripartite classification of mankind. In the first edition of De Generis Humani (1775:41), he had located the second of four geographically defined varieties — comprising 'dark men, with flattened noses' — in southern Asia and the Austral lands and islands. By the second edition (1781:52), the desire to be 'more consonant with nature' saw Blumenbach identify a fifth human variety spread between the island groups immediately beyond mainland Asia, which were inhabited by 'men of a uniformly very dark colour, with broad nose, and thick hair', and 'the Pacific archipelago' which Reinhold Forster had further subdivided into 'two Tribes'. Forster's dual classification of South Sea Islanders was thus firmly inscribed in metropolitan scholarly awareness within three years of its publication. By 1793, in a letter written in English to Banks, Blumenbach naturalized Forster's 'two Tribes' as 'the two principal Races which constitute this remarkable variety in the 5th. part of the world': the 'black race' and the 'brown one'.^[39] By 1795, he had further refined his classification of human varieties and named them 'Caucasian', 'Mongolian', 'Ethiopian', 'American', and 'Malay'. Citing Banks (via Hawkesworth) as his earliest authority, he justified the final name on the linguistic grounds that 'this variety of men' mostly spoke Malay. He did not weight the varieties equally but positioned the Malay as transitional between Caucasian and Ethiopian — between the purportedly original 'medial variety of mankind' and one of the 'two extremes' (see Figure 4, Chapter 1). Far-flung and very diverse, the Malay variety served as Blumenbach's prime illustration of 'insensible transition' within and between his pentad of varieties and supplied the final proof confirming his core argument for assigning 'all the varieties of men thus far known to one and the same species'. In 1799, again referring specifically to the 'Malay race', he summarized his theory of the formation of human races by the 'degeneration' (meaning 'change') of a migrating 'common stock': the degree of deviation from the white 'primitive figure of the intermediate race' towards the extremes was relative to the 'stronger or longer influence of different climates and other causes' on the 'peoples' dispersed around the world. Thus, whereas the 'extreme' form of the Ethiopian race occurred 'under the burning sky of Africa', this race ran into the Malay race in the 'much milder air' of New Holland and the New Hebrides.^[40]

The tension between the rival imperatives of human unity, racial diversity, and the taxonomic impulse is an undercurrent in Blumenbach's discussion of the Malay variety in his 1795 text but it is patent by 1806 in the equivalent section of the second edition of Beyträge zur Naturgeschichte, 'Contributions to Natural History'. In the earlier work, Blumenbach (1795:319-21) used a series of inductive shifts to distil an internal subdivision of the Malay variety from published voyage narratives. He first observed that Malay speakers themselves varied so greatly in 'beauty and other bodily attributes' that the Tahitians had been divided into 'two different stocks': one 'pale' and European-like in facial features, the other very like 'Mulattos'. He cited two authorities in footnotes: the Frenchman Louis-Antoine de Bougainville (1729-1814) who had circumnavigated the globe in 1766‑69 and whose own term 'races' Blumenbach added in parenthesis; and the Portuguese-born Spaniard Quirós who, claimed Blumenbach, had 'carefully distinguished the variety of men inhabiting the Pacific Islands' by saying that some were 'white' while comparing others to 'Mulattos' and others again to 'Ethiopians'. Thus, the second Tahitian 'stock' resembled Islanders further west in the Pacific Ocean, while the inhabitants of the New Hebrides 'gradually' approached the Papuans and New Hollanders, who themselves merged imperceptibly with the 'Ethiopian variety', so that they might 'not unfittingly' be classed with them.

In appropriating voyagers' descriptions of Oceanian people to his classificatory agenda, Blumenbach succumbed to the common historical snare of anachronism, projecting backwards on to earlier representations the seeming realism of his own reifications.^[41] Bougainville's published Voyage — though not his shipboard journal (1977) — did note marked physical differences between two culturally uniform races seen in Tahiti and between Tahitians and 'black men' seen further west. Yet he, like his English translator Reinhold Forster, used race in its multivalent eighteenth-century sense.^[42] Quirós reported a finely discriminated continuum of skin and hair colour and considerable local variety in people he had seen and heard about in both the eastern and western Pacific Islands. However, he did not 'compare' some to the Ethiopians, as Blumenbach thought, deceived by the mistranslation into English by the Scottish hydrographer, Alexander Dalrymple (1737-1808), of Quirós's Spanish adjective loros, 'dark brown', as the noun 'negroes'.^[43] Quirós could not have conceived and Bougainville did not propose a systematic physical typology of Oceanic humanity. Their labels and descriptive terms, though often derogatory and anti-Negro, were conventional, comparative, or experiential rather than regionally categorical (Douglas 2006:10-13). Quirós attributed no specific geographic, racial, or moral coordinates to the human 'differences' he discerned which interested him mainly as supposed signs of the 'vicinity of better governed people' and the occurrence of 'great commerce and communication' — as they would also interest the mid-eighteenth-century compilers of collections of South Sea voyage texts, Brosses and Dalrymple.^[44]

However a priori, the sequential logic enunciated in 1795 shows the morphological criteria of 'analogy and resemblance' — Blumenbach's terms — at work in his taxonomic practice.^[45] Thus far, his occasional, ambiguous usage of the term race and its synonyms tribe, variety, and stock signal the relative lexical insignificance of race, its instability, and its ongoing contemporary genealogical connotations. But in the new edition of Beyträge zur Naturgeschichte revised in the light of Kant's biological conception of races (1806:72), Blumenbach reconfigured 'the black Papus ['Papuans'] at New Holland, etc.' as a Volk whose 'more or less striking formation' distinguished them from the 'brown' Pacific Islanders so that they had become a 'separate' Unterarten, 'subspecies', of the Malay Rasse. There is no such passage in the first edition of this work (1790:83). The reformulation, invoking organic difference more than analogy and nuance, brought Blumenbach nearer to the position recently put by the French geographers Edme Mentelle (1730-1815) and Conrad Malte-Brun (1775-1826) who had pioneered the explicit racial division of Pacific Islanders into copper-coloured 'Polynesians' and black 'Oceanic Negroes'. In the process, they challenged Blumenbach's placement of the New Hollanders within the 'Malay race' on the grounds that he had himself acknowledged their characteristic physical similarity to the 'Ethiopian or African race': his 'system', they sneered, was thus undermined by his own data.^[46] The advent of racial taxonomy with respect to Oceania and the dilution (by Blumenbach) or shelving (by Mentelle and Malte-Brun) of the concept of 'insensible transition' between varieties are textual markers of a hardening in prevailing discourses on human differences and the biologization of the vocabulary available for their description and classification.