National Library of Australia - Home page
NLA News - home February 2000 Volume X Number 5
Lilian Medland (1880-1955) Falco Longipennis and other birds c.1930s, watercolour; 28x19cm, Mathews Collection, National Library of Australia

Gregory M. Mathews donated his library of literature on Australian birds to the National Library as a ‘thank-offering for a successful career’. Richard Schodde explains how Mathews' library is now helping to resolve the chaos he created during his career

The Mathews Collection is a massive, near 5000-piece library covering the literature of Australian birds, from their first discovery in the mid-eighteenth century, to the mid-twentieth century, when Mathews died. Australian-born Gregory McAlister Mathews was its builder between the early 1900s and 1939, when he donated it to the National Library as a ‘thank-offering for a successful career’. He had made sufficient wealth from rural activities, dealing in gold shares and marriage to retire to the life of a country squire in England, indulging what became an enduring hobby: the description and classification of Australian birds.

Gregory M.Mathews
Unknown photographer
Gregory M.Mathews, 1912
photograph; 14.5 x 10 cm
Mathews Collection
From the Manuscript Collection (MS1465)

To feed his mill, Mathews needed information of two types. First, he needed specimens of birds from all corners of Australia, to tell him what kind of bird lived where. Secondly, he needed a complete reference library to tell him which specimens were of kinds of birds already known and which were new to science. His specimen collection became the largest collection of Australian birds at the time—dealing and donations saw it grow to over 40 000 items—but it passed by bizarre circumstances to the American Museum of Natural History in New York after being refused by the Australian Government. The circumstances turned on the second Lord Rothschild of Tring, who sold Mathews' specimen collection with his own monumental collections to cover his losses after being blackmailed over two mistresses. The superb library, too, became the world's most comprehensive library of Australian ornithology of the period. It was Mathews' real pride and joy. He had amassed it over nearly 40 years for the purpose of having at hand every possible reference to arguable questions about the nomenclature and taxonomy of Australian birds.

But what of Mathews' effect on Australian bird study itself? Supported by his growing specimen collection and library, he published his first papers and list of Australian birds in 1907-8, followed by the first part of his monographic The Birds of Australia in 1910. It was the beginning of a monumental 12-volume enterprise—John Gould's The Birds of Australia was only seven volumes—that was not completed until 1923. Yet this was but a fraction of Mathews' prodigious output. Before his death in 1949, he produced another six detailed lists of Australian and Australasian birds, five supplements to The Birds of Australia, over 200 papers and notes in the ornithological journals the Ibis, the Emu, and the Bulletin of the British Ornithologists' Club, and even a journal of his own, the Austral Avian Record. This appeared in five volumes between 1912 and 1923. Almost all the papers in it were published by himself, sometimes with assistance from the English-born New Zealander, Tom Iredale, whom Mathews employed as his amanuensis for the production of The Birds of Australia.

In Mathews' view, his enterprise had created a smooth path for ornithology in Australia: its taxonomy was worked out and its nomenclature settled. But there were contrary views, none blunter than that of C.E. Hellmayr, one of the wisest ornithologists of the early twentieth century, who commented 'it will require a labour of years of conscientious ornithological work to extricate Australian ornithology from the chaos into which Mathews plunged it'.

Banksian Cockatoo
Calyptorhynchus banksii or Banksian Cockatoo
Reproduced from plate 282 in The Birds of Australia, volume VI, by Gregory M.Mathews (London: Witherby & Co., 1916-17)
In The Birds of Australia, Mathews lbaelled plate 282 as Calyptorhyncus macrorhychus or Great-billed Cockatoo

Unfortunately, Hellmayr's grim prediction has proved all too true. But the reasons are complex and, even today, not well understood; and the faults are not Mathews' alone. I shall try to simplify the situation.

I have just used the terms ‘taxonomy’ and ‘nomenclature’ without explaining them. Taxonomy is the science of classification; and nomenclature, for biological purposes, is about putting the correct name to an animal in accordance with its classified position. Together, taxonomy and nomenclature provide the foundation for information on the wildlife of any region on earth. Before you can study any plant or animal, you need to know what beast you are dealing with—its identity and what its correct scientific name is—so you can communicate with others about it. This is what Mathews was into, and through this he wanted to lead Australian ornithology beyond John Gould.

John Gould, the father of Australian ornithology, had compiled the first comprehensive inventories of Australia's birds in the mid-1800s. He classified Australia's birds at species level, then and still the basic classificatory unit of biology. But taxonomic theory had moved on by the turn of the century when Mathews came on the scene. What had changed was that, whereas distinct geographical forms of the one lineage were treated as separate species in Gould's time, they were classified as subspecies or geographical races of one species from Mathews' time on.

Yet these forms, the subspecies and races, in turn became lost to biology through our preoccupation with species. The problem is that regional forms are the building blocks of evolution and the real units of biodiversity in Australia's bird life. Their identification and inventory is a vital cog in the management of Australia's natural resources and it was at this level that Mathews set out to make his mark. He was not driven by the value of his work for conservation; quite simply, he wanted to be the first to document all forms of Australian birds at all levels.

Here he ran off the rails. Due partly to inexperience—he lacked professional training—and to his independence and cupidity in adding to his collection—he would name any number of races after his collectors and their wives in the hope that they would send him more specimens—Mathews used a scatter gun approach, naming new races from any and every region and diagnosing them with the most trivial descriptions. Nowhere was there explicit analysis or transparent rigour. We now know that there are about 1200-1300 distinct races of Australian birds; Mathews named over 4000—and which of these were real and which were not was anyone's guess. That is the taxonomic side to Mathews' contribution to Australian ornithology.

T.H.Maguire (1821-1895)
John Gould, Esq., the Ornithologist
Published in Illustrated London News, 12 June 1852
print of wood engraving; 40.6 x 26.3 cm
Pictorial Collection

Now let us consider nomenclature and the role of Mathews' library. Mathews needed his library to tell him which races were new, which were already known, and what the correct name of each known race was. Determining correct names is an unloved, legalistic, yet crucial task in biology. If carried out in accordance with a set code—and there is such a code ratified by the international biological community—it provides unique and stable names for each species and race. This is the province of international zoological nomenclature, which is founded on four cardinal principles: the name for each organism must be unique; each name must be accompanied by a diagnosis or other indication of its identity, such as an illustration; the name first published for an organism has priority over others; and the name must be tied to an individual marker or ‘type’ in the taxonomic group to which the name applies, as a crosscheck on the application of the name. For species and races, the type is an actual specimen.

Where is the information to be found that reveals whether a name is unique, whether it is properly described, whether it is the first published for a species or race, and what its type is? It is found in reference libraries of taxonomic publications; and the more comprehensive that library, the more accurate and reliable our nomenclatural decisions are likely to be. This gives the first glimmerings of just what a priceless resource the Mathews Collection is for Australian bird study. Except for one surprising gap—Latham's Supplementum Indicis Ornithologici, in which the new species drawn by lay artists of the First Fleet are described—the Mathews Collection is an almost complete reference for the literature of Australian birds up to 1950. And this covers over 95% of the real species and races of Australian birds.

Mathews himself was an inveterate bookworm and nomenclaturist, and almost no-one anywhere dared to cross swords with him on nomenclatural matters concerning Australian birds. If we despair at his splitting of races and genera, we marvel at his great library, which allowed him to root out one knotty nomenclatural problem after another. Yet even here Mathews suffered from lapses of scholarship and intrusions of prejudice such that all his conclusions have to be rechecked to sort the right from the wrong.


not very good, but sufficiently so,” I might explain that the majority of Reichenbach's figures are excellent and recognisable at sight. Again, it is conjectures that Reichenbach took up the name Cyclopsitta from Hombron and Jacquinot's plate of their Cyclopsitte double oeil which appeared in 1846. The plate does not show the details given by Reichenbach, which must therefore have been drwn from a specimen. The species named by Hombron and Jacquinot is the one which I have written up my generic characters, and it differs in every essential, so that Ogilvie-Grant's careful consideration is valueless, as the accompanying figures taken from that bird prove.

The history of the name Opopsitta will now be given.
In an article in the Proc.Zool.Soc. (Lond.), 1860, p. 223 et seq., Sclater, writing on the genus Prioniturus, listed the Parrots of the Eastern Archipelago. On p. 227 he included Opopsitta diophalma and Opopsitta desmaresti. Whatever his intention was canot now be known, and fifty years after he wrote me that it was a slip for Cyclopsitta. That may have been, but in 1862 Rosenberg accepted the genus name Opopsitta, and again in 1873 Sclater himself made use of it. From the published account the name appears valid, as Opo- does not seem to be printer's error for Cyclo-, as Ogilvie-Grant

Mathew's reproduction of Reichenbach's figure of Cyclopsitta or 'Australasian' Fig-parrot
Reproduced from The Birds of Australia, volume VI, by Gregory M.Mathews (London: Witherby & Co., 1916-17) It wasn't until the late 1970s that it was discovered that Mathews had omitted Reichenbach's crucial figure of the lower bill from his reproduction

Here are two examples of Mathews' flawed conclusions. The first concerns the Australasian Fig-parrot, which throughout the late nineteenth century was known by the generic name Cyclopsitta, meaning Cyclops parrot. No doubt it was intended by its describer, the Dresden ornithologist, H.G.L. Reichenbach, as a juicy antonym to its specific name diophthalma, meaning double eyed. Reichenbach first published the generic name Cyclopsitta in association with a set of figures, without description and without nominating any associated species or type. When compiling the Australian parrots for The Birds of Australia in 1916, Mathews reviewed the drawings and concluded that they could not be identified with fig-parrots, or any other parrots for that matter. So he rejected Cyclopsitta, and chose another later name as the generic name for the fig-parrots. To prove his point, he reproduced Reichenbach's figures. The ornithological world took Mathews' conclusion at face value—until the late 1970s when we re-examined the problem by consulting Reichenbach's original publication in the Mathews Collection. There we found that, whatever the reason, Mathews had omitted Reichenbach's crucial figure of the lower bill from his reproduction; it is the one feature in all Reichenbach's figures that is absolutely diagnostic of a fig-parrot, with its three ridges. And so today, the fig-parrots are known once more by the generic name of Cyclopsitta.

Then there is the case of the Great Bowerbird of northern Australia. It is represented by two distinct races, one dark with a plain head in north-western Australia and one paler and scalloped silvery on the crown in north-eastern Australia. The first specimens to reach England in the 1800s, via Alexander Macleay, Colonial Secretary for New South Wales, were then named as the species nuchalis by William Jardine and P.J. Selby in their Illustrations of Ornithology. Although no source locality or type specimens were quoted, the scuttlebutt around the ornithological traps at the time had it that nuchalis came from north-western Australia. This led Gould to describe the north-eastern form as a new species, orientalis.

Gregory McAlister Mathews at Eastlake, 26 June 1940
Unknown photographer
Gregory McAlister Mathews at Eastlake, Sydney, 26 June 1940
Mathews Collection
From the Manuscript Collection (MS1465)

But in 1912 Mathews assessed Jardine and Selby's uninformative description as referring to the north-eastern form instead. This allowed him not only to sink Gould's name orientalis under nuchalis for that form, but also to choose a new name of his own for the now supposedly nameless north-western race.

And, of course, he named it after one of his collectors. It was a ploy that Mathews used time and again in the course of deciphering and cataloguing the names of Australian birds. And this particular case has caused ongoing dissension in Australian ornithology, without any investigation of the root of the problem.

Jardine and Selby's original publication of nuchalis is in the Mathews Collection. To be sure, the description of nuchalis is superficial and uninformative, but fortunately it is accompanied by a coloured illustration. And this figure shows clearly that it is of the north-western, not north-eastern form, with a dark body and plain head. The plate resolves the problem. Did Mathews overlook it? Did he fail to recognise the characteristics of the two races? Or was he fiddling the books? We may never know.

These examples reveal the real strengths of the Mathews Collection in Australian bird study. It contains all the original source material for accurately establishing the scientific nomenclature of Australian birds. It is the reference that makes compilations of checklists and catalogues, such as the bird volumes of the Australian Government's Zoological Catalogue of Australia, not only possible but also reliable as official records of the correct scientific names for Australia's birds. If Mathews created chaos in the taxonomy and nomenclature of Australian birds, it could and is being rectified by research. If his collection of bird skins in the United States is difficult for Australians to consult, it could and is being replaced by the collecting programs of Australian museums. But if he hadn't collected all the rare books and obscure pamphlets and papers covering the original source names and descriptions of Australian birds, we would never have had the literary foundation that we do today for placing the national catalogue of our birds on a sound nomenclatural footing.

DR RICHARD SCHODDE is Curator-in-Charge of the Australian National Wildlife Collection at CSIRO Wildlife and Ecology, Gungahlin, ACT. Trained as a systematic botanist, he has specialised in the taxonomy, evolution and biogeography of Australian birds for over 30 years

Return to February NLA News | Return to NLA News Entry
NLA News - home

To order a particular issue of the NLA News e-mail
or join the Friends of the National Library of Australia and receive a copy each month by mail.