Greg Radick, 2013. “Intellectual Property, Plant Breeding and the Making of Mendelian Genetics,” with Berris Charnley. In the Owning and Disowning Invention special issue. Studies in History and Philosophy of Science 44: 222-33.

1. Tradition and counter-tradition in connecting Mendelism and breeding

“In these pages I have only touched the edge of that new country which is stretching out before us, whence in ten years’ time we shall look back on the present days of our captivity.” So wrote the Cambridge biologist William Bateson in 1902, near the end of Mendel’s principles of heredity: A defence—the first textbook in the science that Bateson, a few years later, would enduringly dub “genetics.” He went on: “The breeder, whether of plants or of animals, no longer trudging in the old paths of tradition, will be second only to the chemist in resource and in foresight.”¹ Not long after, Bateson addressed plant breeders directly about the new science and its practical promise. He was notably less condescending about what breeders knew already about their business. But—as documented in a paper on Bateson elsewhere in this special issue—then as later, he never stopped talking up the transformative potential of Mendelism as an applied science or offering himself as an expert in its applicability.² A Cambridge protégé, Rowland Biffen, was no less confident in his predictions of what Mendelism would do for breeders, and no less committed to making those predictions come true. “Of one thing we may be certain,” Biffen wrote in 1904, at the end of a paper in the Journal of the Royal Agricultural Society of Englandon his Mendelian experiments with wheat and barley. In the near future, “research in the right direction will make clear what now appears as mysterious, as did the results of the breeders who had to work in ignorance of Mendel’s discoveries.”³

The usefulness of Mendelism for breeders—and, relatedly, Mendelism’s role in breaking down barriers between biological theory and breeding practice—has ever after remained a major theme for commentators on the rise of Mendelian genetics. In 1910, an American ally of Bateson’s, Charles Davenport, writing in the first issue of a new “journal of genetics and eugenics” sponsored by the American Breeders Association, noted the passing of “the scholastic biologist of our universities” who little appreciated artificial breeding and looked down on the men involved with it. (He recalled one of these colleagues asking him what he did at Association meetings—“inspect ‘hawgs’, pass around ‘pertaters’ and show up your biggest ears of corn?”)⁴In 1950, Davenport’s former student William E. Castle told of the warm welcome that Mendelism received in America from those “interested in the study of evolution from a pure science viewpoint” and those who “saw in Mendelism a new tool for the production of new and improved varieties of plants and animals.”⁵Nearer our own time, this perspective continues to be well represented in the professional historiography on genetics. Diane Paul and Barbara Kimmelman on the American reception of Mendelism, Robert Olby on the British reception, Kyung-Man Kim on the debate over Mendelism in both countries: all have stressed the importance to the Mendelians’ success of their reaching out across the divide to breeders—and with a theory that illuminated breeders’ problems as none had before.⁶

This version of the rise of Mendelism is, for certain sensibilities, perfumed with an air of “history from below.” Against, say, a complacently elitist view of Mendelism as needing only the backing of the big-question biologists, we learn that it was the breeders (“associated in the mind,” wrote Davenport of the scholastic snobs, “with the cowboy, the stable boy, the ‘hayseed’, the country jay, and the peasant of Europe”), who made all the difference to Mendelism’s success.⁷ Even so, the version’s start in the polemics of partisans means that the notion that breeders were early converts to Mendelism can hardly be treated as unproblematic. Recent historical studies of Mendelism’s international life are already well on their way to building up a more complicated picture of Mendelism’s connections with breeding. Christophe Bonneuil’s work on the situation in France, Jonathan Harwood’s on Germany, and Bert Theunissen’s on the Netherlands have gone some way to querying the dependence on Mendelism of the advances in breeding in those countries in the first half of the twentieth century. For the Anglo-American case too, there have been, over the past quarter century, some important dissents, notably Paolo Palladino on successful breeders who were non-Mendelian or half-heartedly Mendelian, and Richard Lewontin and Jean-Pierre Berlan on whether hybrid corn was the Mendelian wonder-plant of legend.⁸

Where does the truth lie? “Somewhere in between” is the obvious and, in our view, correct answer. But to leave the matter there, in irenic obscurity, is to disregard what is surely the real challenge that these historiographies raise: namely, to show how, exactly, and in spite of the limitations of Mendelism as a guide for practical breeders (as per the counter-tradition), Mendelism came to be accepted as transforming the scientific basis of breeding, in ways that eased the wider acceptance of Mendelism (as per tradition). It is the burden of this paper to take up this challenge. In doing so, we shall make use of the tools developed in the introduction to this special issue for analysing interactions between intellectual property narrowly construed and broadly construed. Our aim is to map out a hitherto unrecognized Mendelism-breeding borderland, occupied by rogues—“rogue” being the name, then as now, for individual plants deviating from varietal type—in order to show, first of all, how intense and consequential were the intellectual-property interactions taking place there, and secondly, how an appreciation of these interactions can lead to a new, composite understanding of what Mendelism did for breeding and vice versa. Although we will be concerned almost exclusively with the British theatre of events, we hope nevertheless to shed light on the “Mendelian revolution” generally.⁹

Our starting point is a look at the functional equivalents of patent claim-making and defending—“IP-narrow” activities, in the program shorthand— in the world of the British plant breeders in the decades around 1900, paying particular attention to the role of rogue plants in ownership disputes, and to the growth of institutions of different kinds for managing, and minimizing, those disputes. Next we provide a counterpart survey of “IP-broad” activities among British biologists of the period involved in theorizing inheritance, where we seek to show how that theorizing took place amidst claim-making in relation to two of the domains of IP-broad discussed in the introduction, productivity (“whose theory, because true, will be most useful?”) and priority (“whose name deserves to attach permanently to the science of inheritance?”). The remainder of the paper draws upon these surveys of IP-narrow and IP-broad to trace out a previously hidden history of how rogue plants figured in the scientific debate over Mendelism, concentrating first on the work of a critic, the Oxford-based W. F. R. Weldon, and then on the work of a convert, Biffen. Although he is not known for it now, Weldon did more than anyone to make conspicuous the nature and extent of the conceptual threat that rogues posed to any Mendelian science of inheritance—and he was able to do so only thanks to the particular IP-narrow arrangements that pertained among British plant breeders. Furthermore, it was changes in these arrangements which, as much as anything else, enabled Biffen and his allies gradually to make persuasive longstanding promises that Mendelism would guide the breeding into being of rogue-free varieties.

2. IP-narrow: Ownership of new plant varieties in Britain in the decades around 1900

“Intellectual property” was not a term that anyone in Britain in the decades around 1900, plant breeder or otherwise, would likely ever have come across. But then as now, plant breeders had problems for which the phrase is apt.¹⁰ Budding British breeders who read the chapter on breeding in John Percival’s textbook Agricultural botany, first published in 1900, were put on their guard. Do not, Percival warned, assume that just because you have bred into being and named a new variety, or have purchased seeds for a supposedly new variety from somebody else, that all is well or will long remain so. First of all there is the problem of rogues—a term Percival put both in quotation marks and italics, and defined as individual plants “departing considerably from the type” and which “appear among the offspring at irregular intervals.” In these departures, he went on, rogue plants “most frequently exhibit characters possessed by the ancestors of the variety in which they are found.” And these unwelcome representatives of the past were bound to turn up, Percival counselled. Indeed, so familiar was the “tendency of plants to revert to long-lost characters” that he reported several labels for it: “atavism, ‘throwing-back,’ or ‘reversion.’” Against reversion, the breeder had but one weapon: the hunting down and destroying of rogues when they appeared. But not everybody could be counted on to be vigilant, and that made for constant trouble:

Very few if any varieties of plants propagated by seeds remain like the type first sent out by the raiser for more than a limited number of years. In a great many instances where almost everybody raises seed, destruction of “rogues” is not efficiently or thoroughly carried out, and through the consequent mixing with the progeny of the reverted plants, the type rapidly degenerates in purity.

Nor was that all. Even the most zealous breeders, working with utmost skill to preserve a new variety, would often contribute nevertheless to its disintegration. Percival used a (fictional) variety of pea as an example:

[T]hree different raisers of seed of “Gubbins’ ‘Incomparable’ pea” are almost certain to hold different views from Mr Gubbins and from each other in regard to the relative importance of the various characters of a good pea; selection is therefore carried out from three different standpoints, and in a few generations the “Incomparable” variety no longer exists except in name, unless Mr Gubbins himself also carries on the propagation: three different types bearing the same name would arise. It is therefore very necessary for the farmer and gardener not to be led away by the fascination of an old name, for it does not follow that anything useful is obtained with it; at the same time it must be remarked that a new name does not necessarily represent any new quality or character in the seeds to which it is applied; new names may easily be applied to old articles when the latter cannot be sold by their original names.¹¹

The vagueness and passive-voice construction of that last warning enabled Percival to pass in silence over the sometimes nasty commercial realities that made these difficulties so much worse in practice. For those other raisers of seed were not just Gubbins’ customers; they were, potentially if not actually, his rivals. And given the success of his variety, it could very well be in their interests—and not at all in his—for them to attempt to sell rogue-riddled or otherwise inferior pea seeds under the “Gubbins’ Incomparable” name; or conversely, to sell his pea seeds under names of their own.

Within the history of biology, 1900 is of course associated not with Gubbins’ peas but with Mendel’s. That was the year when Mendel’s studies in experimental hybridization from decades before suddenly—and for reasons we shall examine below—became a talking point among experimentally inclined botanists interested in heredity. Percival’s textbook, which went off to the publisher in March 1900, made no mention of Mendel. The Brünn monk’s experiments and explanations, clearly and respectfully summarized over ten pages (out of eight hundred plus), show up only in the fourth edition of 1910. Not, it should be noted, in the plant-breeding chapter—where coverage of the role of hybridization in the breeder’s armoury remained unchanged from the 1900 edition—but in the preceding chapter, on reproduction. That placing notwithstanding, Percival saw Mendelism as of more than strictly theoretical interest. After introducing the Mendelian basics, he drew attention to a couple of ways in which, in his words, the “Mendelian conception of distinct unit characters”—the notion that, for example, the colour of a pea seed may be all-yellow or all-green, but not something in between, and that seed colour gets inherited independently from other inheritable characters—“ … greatly assists the efforts of the plant breeder.” First of all, the breeder seeking to combine characters from different plants might draw on Mendelism to plan out a simpler and more direct series of crossings than would otherwise be possible. Furthermore, the new science brought some order to the chaos of rogue plants. “Mendelism,” wrote Percival, “… throws considerable light on various forms of reversion”—notably, those cases where “‘reverted’ individuals … are merely recessives which have never had the chance of showing themselves.” So, for example, the breeder perpetually disappointed in his attempts to breed uniformly yellow-seeded peas but forever finding green-seeded ones among them would, with Mendelism’s help, come to appreciate that the starting, yellow-seeded stock must have been hybrid. In the Mendelian vocabulary, “dominant” characters such as yellow are visible whether the plant contains only yellow-making factors or whether it contains both yellow-making and green-making factors. Given a hybrid ancestor, offspring generations will eventually—and with a regularity which Mendel quantified and explained—produce plants containing only the green-making factors and so displaying greenness (the “recessive” character). Of the return of characters from much further back in a plant’s lineage, however, Mendelism was silent.¹²

In Percival’s perspective, then, the Mendelian conception was helpful to the breeder, but it was hardly revolutionary. The main how-to messages on plant breeding in his textbook—standard reading in applied biology in Britain throughout the first half of the twentieth century—remained the same in 1910, and indeed in 1950, as they had been in 1900.¹³

What of the breeders’ world more widely? “Synonyms” was the term used in connection with the IP problems gestured toward by Percival, referring both to the situation of different plant varieties bearing the same name and, confusingly enough, to the opposite situation of one variety bearing several names. Synonymy was the plant world’s version of piracy. On the whole the market was unregulated, so the sale of synonyms could be quite widespread. Furthermore, the hierarchical structuring of the market meant that, although less established breeders could be “discouraged” more or less effectively from the practice, better established breeders sold synonyms with relative impunity.¹⁴ But even in these cases, there were various informal means of protection; and there gradually emerged an increasingly organized set of instruments and even legislation to deal with synonyms and other problems such as deliberate mislabelling and the sale of inferior seed. We shall first describe some informal practices which persisted throughout the period, then look at some changes in formal structures.

In Britain at the time there were several big seed firms, including Garton’s, James Carter & Co., and Sutton’s. The literature produced by these companies, including seed catalogues and farmers’ guides, played an important role in protecting new varieties from piracy. Carters, in particular, produced lavish yearly catalogues and farmers’ guides, often with photographs that sometimes even showed how different varieties compared with each other in order to aid identification. In general, the reputations of these companies served as a guarantee and a means to keep buyers coming back to the same source of seed; this promotional literature was one way of maintaining these hard-won reputations. Carters were particularly keen to trade on their name and position as suppliers of seed to the royal household—a distinction they proclaimed in much of their advertising literature. Another set of means for securing identity, and so protection, was the use of sacks and seals. Seed was often sold in sacks with the name of the seed variety written on them, and sometimes these sacks were also sealed (see Fig. 1 for an example of a seal used by Carters). The sealed sacks would then be supplied directly to the purchaser by post, further reducing opportunities for tampering.

 

Fig. 1. Seal for seed sacks used by Carters, ca 1900. From a letter to Weldon from Carters, 2 Dec. 1901, in the Papers of Karl Pearson, 264/2, Special Collections, University College London.

Even in the nineteenth century there was an important institutional dimension. Informally, the Royal Horticultural Society (RHS) provided some protection against synonyms. The Society ran several committees which rated new varieties, awarding certificates based on quality: first class, second class, botanical commendation and commendation. Awards of these certificates were reported in the Society’s journal. Notice of awards would often be made in firms’ advertising material, where they functioned as a warrant to the originality and quality of a new variety (see Fig. 2 for an example of Carters’ advertising, illustrating the company’s use of prizes). These certificates were augmented by the awarding of prizes at flower shows (the most famous of which, the Chelsea flower show, is still running). Another institution, the Royal Agricultural Society, also ran some trials of new varieties, publishing the results in its journal. Readers would often write in with the results of their own trials. The Royal Agricultural Society’s shows also awarded prizes and so credit in a similar way to the RHS. Independent journals such as the Gardeners’ Chronicle—which enjoyed a similar circulation to the Guardian and the Economist in the period, and is well known to historians of biology for contributions from the likes of Darwin and Hooker—provided forums for the promotion of new varieties and, through their correspondence pages, the airing of disputes as to originality.

 

Fig. 2. Advertisement for Telephone and Telegraph peas taken from Carters’ Illustrated vade mecum and seed catalogue (1879). The catalogue was also intended to function as a handbook, a vade mecum, which translates literally as “go with me.” Image supplied by the RHS, Lindley Library. © RHS.

Thanks to these and related institutions and publications, seed dealers accrued reputations around particular varieties. To a large extent it was this informally established and maintained reputation, coupled with direct sales, which protected the largest, most established dealers, at least, against the sale of their varieties as synonyms. Nevertheless, there was also, as noted, some development over the period of more formal means for protecting the identity of varieties. The oldest piece of legislation was the Adulteration of Seeds Act of 1869. This Act dealt mainly with outlawing various practices used to make old or bad seed saleable. But the Act was largely seen as toothless, since there was no official body to enforce it. Not a great deal more happened until the twentieth century, in particular the period after the Great War. The disruption to the agricultural status quo during the war concentrated minds on the question of how government could best help agriculture, especially as the threat of isolation from trading lines had now become a very real prospect. One of the results of this shift in official attitude was the temporary Testing of Seeds Order of 1917, instigated as a special measure during a part of the war when all agriculture was under government control. The Order stated that all seeds should be certified for identity, germination level and purity, understood as freedom from weeds and disease.

After the war this legislation in turn became the basis for the Seed Adulteration Act of 1920, which demanded the use of certificates for all seeds, produced at point of sale. The government Ministry for Agriculture and Fisheries—a forerunner of the current Department of Environment, Food and Rural Affairs (DEFRA)—oversaw enforcement and provided inspectors to take samples from thousands of businesses that sold seeds, including farms that sold to other farms, and even blacksmiths or grocery stores that sold seed only seasonally and in very small quantities. Enforcing the Seed Adulteration Act in turn generated a demand for seed testing services. A newly established charity, the National Institute of Agricultural Botany (NIAB), based in Cambridge, became the new home of the chief English seed testing station, the Official Seed Testing Station (OSTS). The OSTS at NIAB was responsible for checking the particulars given in the seed certificates and providing those particulars to vendors. NIAB also published yearly reports on the OSTS’s testing activities in its own journal. NIAB also had another regulatory role: the testing for quality and distribution of new varieties raised by publicly funded research. The seeds for these varieties would be sold by NIAB, which acted as intermediary between research centres and the established seed corn dealers. In this way NIAB exerted some control over the sale of seeds while at the same time utilizing established supply channels and advertising of seed companies such as Carters, Sutton’s, and Garton’s. The Institute was also, significantly, responsible for bestowing credit on the new varieties it tested and then reciprocally harvesting that reputation to bolster its own.¹⁵

A crucial but easily overlooked additional innovation of relevance here was the growth, alongside publicly funded agricultural research and seed testing, of an ideal of selfless public service amongst the researchers who depended on that funding. For a glimpse of that ideal in action, consider the following extract from a speech made in 1924, introducing one of the most famous of the publicly funded breeders, Rowland Biffen. The speechmaker was Sir Herbert Trustram Eve KBE, introducing Biffen at the London Farmers’ Club:

We practical business men, if we have an idea, try to make money out of it; it is human nature, but the scientific man is always working for others without advantage to himself [ … ] There is no patent, there is no copyright in seeds, and yet our scientific friends are spending the whole of their lives in seeing how they can help the farmers of this country.¹⁶

3. IP-broad: Ownership of the science of heredity in Britain (and elsewhere) in the decades around 1900

The accomplishments that, by 1924, made Biffen so esteemed will be dealt with in more detail below. In attempting to take their measure fully we shall draw upon the IP-narrow background surveyed in the previous section. But we shall need too to make appeal to some IP-broad background, including considerations to do with Biffen’s own hero, Gregor Mendel.

Of the paper for which Mendel is still so famous, Biffen wrote that, “judging from the almost absolute lack of reference to it by later writers, it was completely lost sight of.”¹⁷ That was an exaggeration, though a persistent one. In the later nineteenth century, Mendel was not quite the forgotten figure of legend. Botanists had regularly and respectfully cited his pea-hybridization paper, published in 1866. But Mendel had taken his place among several investigators who were considered to be working along similar lines. Where he had identified and, in a provisional way, explained patterns of inheritance for particular traits in peas, others had done roughly the same for other plants. No one attributed much significance to the “law valid for Pisum,” as Mendel had called it.¹⁸ And then, in the spring of 1900, the German botanist Carl Correns published a paper enshrining “Mendel’s law” in its title. As Correns explained, his own researches with maize as well as peas had led him to rediscover what, he had belatedly realized—and, he noted, another rediscoverer, the Dutch botanist Hugo De Vries, had apparently yet to realize—Mendel had discovered so long before.¹⁹ Within a few years, Mendel’s law or, more commonly, laws came to be hailed, in influential quarters, as the basis for a new and general science of inheritance. Soon rebranded “genetics” or “Mendelian genetics,” it was initially called simply “Mendelism.” Over a century later, in our own genomics age, “Mendelian inheritance” is still a meaningful phrase, and Mendel remains the celebrated founder. In the words of a popular undergraduate biology textbook, his “theory of inheritance, first discovered in garden peas, is equally valid for figs, flies, fish, birds and human beings.”²⁰

And yet, in the 1900 paper which singled out Mendel’s law so consequentially, Correns made rather modest claims on its behalf. He guessed that it would probably turn out to be valid only for “varietal hybrids”—that is, hybrids formed from varieties within a species, as distinct from hybrids formed from different species—and then only for those varietal hybrids where one character in a pair dominated the other. Even some hybrid peas, he continued, were a poor fit.²¹ Why, then, make such a fuss about Mendel’s having priority as discoverer? As the sociologist Augustine Brannigan observed in a classic discussion of the Mendelian “rediscovery,” Correns’ generous gesture toward Mendel was at the same time a bid to undermine De Vries. Correns might have lost out to De Vries in the publication race; but now, in stressing how much both men’s work shared with Mendel’s, down to the “strange coincidence,” in Correns’ phrase, of De Vries’ replicating the abbot’s vocabulary of “dominating” and “recessive,” Correns got his revenge. If he would get no credit for the discovery, neither would De Vries. (Intriguingly, both the word and the quotation marks around “rediscovery” are Correns’. Even so, he, like De Vries, owed a larger and earlier intellectual debt to the 1866 paper than he would ever admit.) So Mendel entered the wider biological consciousness, at the time that he did, as a means to the end of resolving a priority dispute.²²

In its very name, then, Mendelian genetics is emblematic of the significance of intellectual property (in our broad sense) in the sciences and the controversies surrounding it. And the name rapidly came to stand for far more than merely that science of heredity which took as its starting point the patterns and explanations found in Mendel’s paper. Thanks above all to William Bateson, “Mendelian genetics” became the name of the science of heredity tout court, with Mendel represented as the trailblazing bringer of scientific method to the study of inheritance.²³

With Mendel’s promotion in this guise came, needless to say, the demotion of other figures whose work could be seen as foundational, above all the English mathematical polymath Francis Galton. An exact contemporary of Mendel’s (both were born in 1822), Galton made his debut as a scientific student of heredity with a pair of articles on “Hereditary talent and character” published in 1865, the same year that Mendel presented his pea experiments before the underwhelmed Brünn Natural History Society. From then until his death in 1911, Galton did more than anyone, anywhere, to attract attention and ambition to the problems of heredity as full of intellectual fascination and social utility. He introduced new and enduring methods of attack, including the analysis of pedigrees and studies of twins. He popularized the phrase “nature and nurture” (in the title of an 1874 book on English scientific men) and coined the term “eugenics” (1883). He developed a physiological theory of inheritance which pictured individuals as harbouring hereditary elements some of which had visible effects (“patent” elements) and some of which did not (“latent” ones). His 1889 book Natural inheritance was a showcase for his quantitative approach and the statistical concepts he had invented in developing it, notably correlation and regression. His readers included Bateson and his friend and fellow zoologist W. F. R. Weldon, young men who, over the next decade, gradually abandoned the evolutionary morphology in which they had been trained for Galtonian kinds of research and mentoring alliances with the man himself.²⁴

Both Bateson and Weldon were enormously impressed with Galton’s proposal, in the late 1890s, of a new scientific law governing inheritance, as disclosed in an analysis of data from the breeding of basset hounds. There was never any consensus about exactly what was governed by the law—which came to be known as the “law of ancestral heredity”—or how far it applied strictly. But the core idea was that hereditary influence could be thought of as dropping away regularly as if in a mathematical series, with parents accounting for one half of the offsprings’ character, grandparents for one quarter, great-grandparents for one-eighth, and so on. That idea seemed to its admirers well suited for handling, in the language of the day, atavisms, reversions, throwbacks; for if ancestral influence diminishes over time without ever going fully extinct, then vanished forms and characters are expected to return, fully or in part. Such returns come as variably and yet as regularly as they do, on the Galtonian view, not because ancestry exerts some sort of force on the future, but simply because, as different combinations of internal and external circumstances chance to arise, what has been latent is bound to become patent again.²⁵

Bateson eventually came to think that Galton’s law represented a special case of Mendel’s laws, with Weldon coming to very much the opposite conclusion. We shall say more below about Weldon, Mendel’s laws and Galton’s law. For now, the important point apropos of IP-broad is to notice how, in pressing their claims for Mendel as founder (a priority claim) and for his discoveries as useful because true (a productivity claim), the Mendelians effectively wrote Galton out of the history of their science. In the 1909 edition of Mendelism, a popular textbook written by Bateson’s Cambridge colleague Reginald Punnett, a chapter entitled “Applied heredity” (published separately in Harper’s Monthly Magazine in December 1908) proclaimed that “the great and baffling problem of heredity has suddenly passed from the speculative to the experimental stage. The credit of it belongs to one man.” There followed a capsule biography of Mendel and summaries of his principles and their recent applications, in plant and animal breeding—Biffen’s achievements as a Mendelian breeder of superior wheat varieties received extensive coverage—and in human heredity, above all as it touched on health and disease. “Professor Biffen’s classic experiments with wheat rust have opened up a fascinating field of research in connection with problems of immunity,” wrote Punnett in conclusion:

… We must have full and accurate pedigrees, and for their interpretation we require carefully devised experiments in the breeding of plants and animals. With increase in knowledge will come powers of prevention far greater than those we have to-day. How far we may use these powers must rest with the future to decide.²⁶

There is no mention of Galton. How striking to find Galton himself, in the conclusion of an address published that same year, sounding much the same note in relation to recent, but rather different, discoveries about heredity:

All I dare hope to effect by this lecture is to prove that in seeking for the improvement of the race we aim at what is apparently possible to accomplish, and that we are justified in following every path in a resolute and hopeful spirit that seems to lead towards that end. The magnitude of the inquiry is enormous, but its object is one of the highest man can accomplish. The faculties of future generations will necessarily be distributed according to laws of heredity, whose statistical effects are no longer vague, for they are measured and expressed in formulae. We cannot doubt the existence of a great power ready to hand and capable of being directed with vast benefit as soon as we shall have learnt to understand and to apply it.²⁷

Our biology textbooks, so solicitous of Mendel’s achievement, tend, like Punnett, to silence over Galton’s achievement. That would have surprised and disappointed Weldon no end.²⁸

4. Mendelism under attack: W. F. R. Weldon, Telephone, and rogue plants as a problem for breeding practice and Mendelian theory

On conventional tellings of the rise of Mendelism, one would never guess that Weldon’s critique of the Mendelian perspective, expressed most pungently in a 1902 paper in Biometrika, the house journal for Galtonian biology, had anything to do with plant breeders’ struggles over ownership or with Mendelian pretensions to supply breeders with useful truths. The Oxford-based Weldon, and the other “ancestrians” or “biometricians” who championed Galton’s law over Mendel’s laws (including, most notably, the UCL mathematician Karl Pearson), have come to be remembered as exemplifying the academic-snob style of biological inquiry lamented by Davenport as defining the pre-Mendelian era. Indeed, Weldon’s alleged indifference to the breeders and their needs has now become an item on a fairly standard list of ways in which, over the course of the 1890s, he and Bateson became ever more divergent scientifically, with Bateson embracing not just breeders but experimental methods and saltational evolution, and Weldon rejecting all of them, at the cost of Bateson’s friendship. The list makes two features of what is now known as the “biometrician-Mendelian debate” that ran from 1902 (when Weldon published his paper) until 1906 (when he died unexpectedly) look prefigured: first, its notorious animosity; second, a close alignment between Mendelians and breeders.²⁹

But even the briefest reading of Weldon’s paper makes untenable an image of him as distanced from the breeders’ world. On the contrary, “Mendel’s laws of alternative inheritance in peas” (to give the paper its title) is through and through the work of someone in the thick of respectful reading of, and correspondence with, breeders, above all those whose business was peas. It was on their testimony, and the evidence that some of them sent to him for his own inspection, that Weldon founded his case.

After first setting out Mendel’s data and the laws formulated on their basis (as well as the chances of Mendel getting such close agreement between theory and observations—a calculation later credited to R. A. Fisher), Weldon turned to consider how fully other breeders’ results with peas bore out Mendel’s conclusions about dominance. His answer was: not at all. None had found that, universally, yellowness is dominant to greenness, roundness to wrinkledness, and so on. Nor did anyone else find the strict unit-category pairs that Mendel described. All sorts of intermediates could be found between the colour and seed-shape extremes, as Biometrika readers could see for themselves in two photographic plates included with the paper.

A related accusation that Weldon laid against Mendel was that his perspective exaggerated grossly the ease with which cleanly differentiated new varieties could be bred from old ones. What Weldon had in mind here was the Mendelian contention that ancestral influence can go to zero in a single generation: a massive violation of Galton’s law. Although this contention is not familiar as a key Mendelian principle nowadays, it is easy enough to discern its importance if we briefly return to the textbook yellow-and-green-peas Mendelian cross and consider the green peas that, according to the texbooks, appear in the generation that follows on from a hybrid, all-yellow generation, alongside yellow peas in the ratio of 3 yellow to 1 green. The question for Weldon was: are these green peas (known at the time as the ‘‘extracted recessives’’) identical in hereditary constitution to their green grandparents, despite having had yellow parents? Putting the same question another way, should we expect the descendent greens to harbour no yellow-making factors whatsoever, and thus to show no hereditary influence at all from their yellow parents? The Mendelians answered ‘‘yes,’’ in defiance both of Galton’s law and, as Weldon’s marshalled evidence was meant to show, of the facts familiar to plant breeders, who knew how hard it was to purify away even quite distant ancestral influence. (When Percival, some years later, covered Mendelism in his textbook, he wrote, of extracted recessives: ‘‘Such ‘reverted’ individuals ought to breed true when crossed among themselves or self-fertilised, and this is sometimes the case.’’)³⁰

To illustrate the point, Weldon referred back to a series of letters published in the Gardeners’ Chronicle at the end of the 1870s concerning a quarrel over ownership between two British breeders. At issue was the identity of Telephone, a putatively new, putatively green-and-wrinkled pea variety. Telephone had been released by Carters, who claimed to have produced it by selection from an older variety, Telegraph, bred by a Yorkshire breeder named William Culverwell; and in 1878 the Royal Horticultural Society had duly certified Telephone as a new variety. The trouble arose the following year when Culverwell charged that Telephone was not a wholly distinct and different variety at all, but merely the imperfectly isolated wrinkled-pea stock from Telegraph, which gave both round and wrinkled peas. Culverwell had sold to Carters the whole of the Telegraph stock—and so, in effect, the rights over it, in a situation reminiscent of what was happening in America—but he felt that their isolating the wrinkled peas from Telegraph would ultimately detract from the stock, since the wrinkled peas, which tended to be sweeter, were reckoned to be more desirable than the round ones.³¹ In this way Telegraph would eventually become an inferior sample of the same variety. Culverwell felt that, if this were to happen, his reputation, as the originator of Telegraph, would diminish as the quality of Telegraphdiminished.

For Culverwell, then, it was above all his reputation as a breeder that was at stake. Be that as it may, the possibility that Telephone might not be distinctive enough from its ancestor to merit a new name could be tested; and at Culverwell’s and the editors’ behest, various contributors to the Chronicle grew the pea varieties together. Finally the Chronicle published its verdict on the case: Culverwell was in the right; for, given its demonstrated tendency to produce rounded and roundish peas alongside the advertised wrinkled peas, Telephone was not distinctively different from the stock of Telegraph.³²

What mattered to Weldon was less this conclusion, however, than the fact that controversies like this one occurred at all. Protracted disagreement over the distinctiveness of “recessive”-based breeds was, Weldon argued, predictable and explicable on the Galtonian understanding of heredity, but a surprising mystery on the Mendelian understanding.³³ Applied to breeding, Galton’s law taught that, in fixing a new variety, breeders should always expect an uphill battle in keeping out unwanted characters—or in other, nineteenth-century words, they should expect rogue reversions to ancestral characters. Even in the early twentieth century, Weldon pointed out, twenty-five generations since the originating cross, Telephone seeds remained stubbornly variable, in shape but also in colour, occasionally exhibiting both of the characters in the Mendelian contrast pairs and even intermediate characters.³⁴

In returning to old numbers of the Gardeners’ Chronicle and a dispute over breeders’ ownership of pea varieties—a dispute that left traces for Weldon to find, thanks to processes institutionalized by breeders in order to manage disputes of this kind—Weldon saw himself as documenting the reality of the long reach of ancestral influence, as against Mendelian teachings. The unit characters described as distinct and segregating by Mendel seemed to be anything but in Telegraph/Telephone.

In an audacious closing section, Weldon credited a plant breeder, indeed a highly successful breeder of peas, with having understood all that Weldon was saying decades before. As Weldon explained, in the very year, 1866, that Mendel published his deeply misleading paper on pea hybridization, a much better paper, far more in accord with the experience of pretty much every other observer, was published in English, under the title “Observations on the varieties effected by crossing in the colour and character of the seed of peas,” in the Report of the International Horticultural Exhibition and Botanical Congress, by one Thomas Laxton.³⁵ Weldon gave a preview to his biometrical ally Pearson in a letter of 21 November 1901:

While Mendel was making his “laws,” Laxton, of whom Darwin speaks so often! was crossing peas and making all the main races we now eat. In 1866 he published his impressions of the result of crossing … He says that the peas directly resulting from hybridization “are sometimes all intermediate, sometimes represent either or both parents in shape or colour, and sometimes both colours and characters, with their intermediates, appear.”³⁶

Born in 1830, dead in 1893, Thomas Laxton was one of the most successful nurserymen of the era, with a number of varieties of peas, beans and strawberries to his name, as well as a collaboration-by-correspondence with Charles Darwin on pea hybridization, memorialized in the pages of Variation of animals and plants under domestication. Laxton has rather disappeared from the recent historiography, although H. F. Roberts, in his estimable Plant hybridization before Mendel (1929), devoted seven pages to Laxton’s hybridization experiments.³⁷ In the brief paper cited by Weldon (who quoted from it at length), Laxton wrote about the third and fourth hybrid generations as well; by the time of the latter, he wrote, we find “the seed often reverting partly to the colour and character of its ancestors of the first generation, partly partaking of the various intermediate colours and characters, and partly sporting quite away from any of its ancestry.”³⁸

The upshot was that, on Weldon’s reading, Laxtonian breeders did not expect new varieties, especially those formed through hybridization, to settle down into a uniform character at all quickly. Rather they relied, in Laxton’s phrase, on “careful and continuous selection” to transform the novelties turned up via hybridization into new and stable breeds.³⁹ And indeed, such was in summary the counsel on plant breeding offered in John Percival’s textbook, in 1902 and beyond.⁴⁰

5. Mendelism defended: Rowland Biffen, Yeoman II, and the protection of Mendelian products and principles against rogues

Weldon’s paper infuriated Bateson, who, within a matter of months, wrote and published the Defence whose prophecy about Mendelism and breeding we quoted at the start. Included in its pages was, needless to say, a vigorously contrary reading of Laxton’s lessons.⁴¹ But perhaps Bateson’s most sustained answer to Weldon took the form of his support for the Mendelian breeding efforts of Rowland Biffen.⁴² An early star of Bateson’s Cambridge school, Biffen is still remembered among geneticists as the man who first successfully applied Mendelian principles to agriculture, producing new and productive wheat varieties including Little Joss (1910) and Yeoman (1916) as well as the variety that will be our main concern here, Yeoman II.⁴³ Beyond his Mendelian allegiances, however, Biffen is useful to think with, in that he represented a very different type of breeder from either Culverwell or James Carter and colleagues. Holder of a chair in Agricultural Botany at Cambridge from 1908 and the directorship of the Plant Breeding Institute (PBI) in Cambridge from 1912, Biffen was one of the new academic scientists that characterized Edwardian Britain, who used public monies to advance agricultural research while at the same time using agricultural research to pull in public monies.

Biffen’s agility in positioning himself, and his Mendelian breeding program, at the centre of the emerging academic-state agricultural complex was one element in the success of that program. So was his career-long devotion to making it a success. Consider again his 1904 paper. En route to his Mendelism-promoting conclusion, Biffen reported the results of trials at Cambridge University’s Experimental Farm using two commercially important cereals, wheat and barley, and aiming to “test the possibilities of the application of Mendel’s discoveries.”⁴⁴ In prospect, Biffen predicted, was the fixation of new varieties far more rapidly than in the bad old days of reliance on selection:

All the evidence which has accumulated … goes to show that the characters of the plant or animal are distributed among the sex-cells according to a definite system, and the possible combinations can be foretold with considerable accuracy. To the breeder the value of this knowledge can hardly be estimated. Once he knows the behaviour of particular characters of the varieties he is working with, he can definitely choose the parents which will give him the combination he desires, and obtain it, fixed, in the first or at the latest in the second generation from the cross-bred. This is worth comparing with one’s expectations in the dark pre-Mendelian days. Then one might by chance find the required type among the mixture resulting from the cross-breds; more often it was a case of the selection we hear so much of – the picking out of such a form as the rough-chaffed red wheat which in the following generation might breed true, or with far greater probability (the chances can be easily calculated) would break up into a number of forms similar to those from which it was originally chosen. A further selection from the mass would in all probability give the same result. Small wonder is it that competent breeders have given up as hopeless problems the solution of which we now know to be simple.⁴⁵

Biffen, by then, had ample reason to feel confident. His work on the Experimental Farm had already disclosed, to his satisfaction at least, that a range of characters in wheat came in Mendelian dominance-recessive pairs. When, drawing on that work, Bateson had reported to American breeders in New York, in the autumn of 1902, that beardlessness in wheat was a dominant character and, therefore, that beardless wheat plants might contain profit-wrecking factors for the recessive character of beardedness, his audience treated the news as wondrous (see Radick’s paper, this issue).

The signs from the Experimental Farm remained encouraging throughout Mendelism’s first decade. In the discussion of the applied-heredity horizon in Punnett’s 1909 edition of Mendelism, readers learned that the most exciting of Biffen’s recent additions to the inventory of Mendelian character-pairs in wheat were susceptibility and resistance to rust, a commercially troublesome disease brought on by a fungal parasite. With rust immunity thus revealed to be, in Punnett’s words, “within the control of the breeder to combine with other characters according as he pleased,” Biffen was in the midst of doing just that:

From the knowledge gained through his experiments Professor Biffen has been able to build up wheats combining the large yield and excellent straw of the best English varieties with the strength of the foreign grain, and at the same time quite immune to yellow rust. During the present year several acres of such wheat coming true to type were grown on the Cambridge University Experimental Farm, and when the quantity is sufficient to be put upon the market there is no reason to doubt its exerting a considerable influence on the agricultural outlook.⁴⁶

Yet the old difficulties of fixing new varieties, so that they would reliably “come true to type,” remained an ever-present problem for Biffen. Yeoman, the most famous product of his Mendelian labours after the rust-resistant Little Joss, turned out to suffer disastrously from a rogue problem. When grown by the thousands in a field, a number of out-of-type individuals regularly became obvious, often because they grew taller than the rest. Although his critics suggested that the presence of rogues in fields of Yeoman pointed towards a reversion of the strain to ancestral type, Biffen viewed that possibility as a part of hereditary folklore—repudiated a generation before, with Galton’s law and the biometrical opposition to Mendelism generally. For Biffen, the stable character of his new strains was guaranteed by the Mendelian principles by which they had been generated.

In defending this view and simultaneously defending against the possibility that Yeoman’s problem with rogues testified to a problem with Mendelism, Biffen offered an ingenious explanation. The real culprit, according to Biffen, was the threshing machine, used to separate the corn from the ear, and to separate both from the straw. At the time, threshing machines were often transported from farm to farm, as any individual farmer was unlikely to be able to afford one. In the process of threshing, some corn would become lodged in the machine, which would then travel to the next farm, where, Biffen alleged, the contaminant corn would become mixed with corn intended for planting the following season.⁴⁷

While the Yeoman rogues were not particularly troubling to farmers—financially, the problem was insignificant—to Biffen, the plant’s identity, as a plant of a certain stable character, was crucial, not least because it was so intimately bound up with his own reputation as the great pioneer of Mendelian breeding. Biffen’s solution to Yeoman’s rogue problem was basically to start again. In November 1922, at the first annual general meeting of NIAB, while giving a lecture in his role as Chief Scientific Advisor to the Institute (the most recent arrival in Biffen’s professional imperium), Biffen made his first public mention of the new form of wheat that he had passed to NIAB for testing and, if deemed successful, distribution. Especially striking is Biffen’s dismissal of reversions as a serious problem for the breeder. The report of the lecture virtually opens with the claim that, “There is no difficulty in fixing these types; so-called cases of reversion are traceable to mixture of stocks in travelling threshing machines.”⁴⁸ The main reason he gave for the release of Yeoman II was, in line with the new exculpatory explanation, to purify contaminated seed stocks. A second reason given was that Yeoman II was supposedly a superior variety of wheat. Biffen was quoted in Nature, in 1923, as saying, “the sooner Yeoman is off the market the better.”⁴⁹

In the Journal of the Ministry of Agriculture in September 1924, Biffen again stated that the new strain was a remedy for the impurities of old stocks. At the end of the article, which announced the release of the new variety, Biffen laid claim to the most obvious—and, as we have seen (recall how Carters did these things in the nineteenth century), traditional—form of protection: the seal to be placed on the sacks in which the seed would be sold. “The attention of farmers is particularly drawn to the fact that genuine seed of Yeoman II can only be obtained in sacks closed with the seal of the National Institute of Agricultural Botany.”⁵⁰ These seals were the means of protecting the release of Yeoman II. Tenders were only to be made to NIAB, the price was fixed, and the seed certified as genuine and superior by the NIAB seal on the sacks it was sold in (see Fig. 3 for an illustration of this seal).

 

Fig. 3. The seal used by NIAB to close bags of the new Yeoman II wheat. Image reproduced from Rowland Biffen, “The new wheat Yeoman II”, Journal of the Ministry of Agriculture, 31 (1924), pp. 509–512.

6. How intellectual property matters for the historiography of genetics: Revisions in prospect

Here we have sketched the outline of an IP-inflected history of the making of Mendelian genetics. At the beginning of Mendelism’s rise, institutional arrangements in place within the British plant breeding community to manage controversy over new varieties and their ownership provided the fiercest of Mendelism’s early critics, Weldon, with data used in the 1902 article that provoked Bateson to his book-length Defence (and so, arguably, to advocacy of Mendelism as a full-time, lifelong job). Drawing especially on the record of nineteenth-century controversies over the pea variety Telephone, Weldon contended that rogue plants, of the sort that regularly led to property disputes among breeders, were a fact of the breeding life which Mendelian principles utterly failed to account for. And, on the other side of Mendelism’s rise, it was changes in these very arrangements—changes in which the Mendelian breeder Biffen played a key role—that helped secure the reputation of those principles as true and, connectedly, of the varieties credited to those principles, notably Biffen’s wheat variety Yeoman II, introduced in the early 1920s, as rogue-free and otherwise excellent.

We noted at the outset our hope that, for all the limitations of focus here, the perspective offered—and in particular, the new mapping of the rogue-infested borderland between Mendelism and breeding—will be suggestive for those whose historical interests lie beyond British biology and plant breeding in the age of Weldon and Biffen. An obvious next-step along these lines, it seems to us, would be to extend the inquiry here internationally. Did rogue plants matter as much, or in the same ways, in turn-of-the-century breeding and biological communities beyond Britain? Why or why not? Answers cannot fail to be instructive, about those communities and their relations but also about why things went as they did in the British case, and how much weight to give to events there in understanding the global career of Mendelian genetics. (For a glimpse of how quickly some American breeders became convinced that Mendelism had solved their rogue problems, see Radick’s paper elsewhere in this issue. And for a look at how Biffen capitalized on his success with those problems generally to push his Mendelian imperium into British-controlled Africa, see the paper by Charnley at the end of the issue.)

Even for the British story, however, there is scope for rethinking and research that could well have broader repercussions. Consider—to take but one of many commonplaces ripe for the plucking—the notion that the eventual “synthesis” of Mendelism and biometry shows that there was no empirical or conceptual substance to that debate, just divergent methodologies and ideologies. To be sure, Weldon’s stress on the need to take rogue plants seriously was of a piece with his allegiances to, among other things, Darwin’s theory of natural selection (which, to be an important evolutionary force, requires the existence of copious, small variations), Galton’s law of ancestral heredity, and Pearson’s positivist concern that scientists had to deal with all of the observations, not just those which fitted overly tidy idealizations. What is more, Weldon’s and Pearson’s biometrical approach is rightly associated with quantitative characters such as height, whose distribution in populations takes the graphical shape of bell curves; and from early days it was understood that Mendelian explanations could be stretched to account for the inheritance of such characters, though Mendel had developed his concepts in order to explain patterns of inheritance in qualitative, either/or characters. This was a synthesis; but—as is not sufficiently brought out nowadays—it was on Mendelian terms, with Mendelian concepts constituting the general principles and biometrical topics relegated to the status of the special case. As we have seen, Weldon wanted something different, and his rogue-based critique is a marvelous vantage point from which to look afresh now at Mendelism’s achievements. From a Weldonian perspective, for example, the regular ratios from which all else about inheritance followed for Mendel were not, as Mendel seemed to think, and as Mendelians thought ever after, patterns which were somehow specially revealing about the ways of inheritance as they truly are. They were simply patterns, identified in advance as desirable and then realized thanks to methods which, by ruthlessly excluding any hereditary factors that would mess things up, ensured those patterns would appear. If one designed one’s experiments correctly, one could, Weldon suggested, end up showing that a given character is dominant, or recessive, or neither.

Here is a point of view never encountered in the standard historiography. Its recovery can help us ask, among other worthwhile questions, whether the congruence between Weldon’s critique and aspects of present-day biology might help with curriculum reform in biology teaching, which has remained stubbornly Mendelian in its orientation.⁵¹ Nor need we restrict the rogue-inspired questioning to the biological side of the breeders-biologists relationship. We have seen that, by 1924, variability in British wheat breeds no longer—as it would have for the likes of Weldon—pointed up an anti-Mendelian conceptual lesson. Breeders came increasingly to see rogues in their fields as most likely due to contamination. The sacks and seals used to protect plant varietal identity, used as they had been for decades to protect the identities of non-Mendelian varieties, came to take on a new significance in the era of genetics. Apparent reversion to beyond-the-parents plants or character instability could henceforth be presumptively blamed on “external” contamination, and external contamination alone. Rogues continued to be found, and their causes to be debated; but with time their conceptual significance was lost.

So, at any rate, the situation appears to us, in the light of the reconstruction ventured here. But we expect that, should it one day emerge, a more comprehensive history of British plant breeders and their rogue problems, going beyond 1924 and so embracing the enormous changes that followed in both genetics and the governance of varietal ownership, will bring to light fascinating subtleties and complexities which in turn will enrich historical writing about genetics and breeding, in Britain and beyond. “Treasure your exceptions!” counselled William Bateson (his most famous one-liner). On Weldon’s reading, Bateson did no such thing. But for historians, his advice, when it comes to rogues, remains exceptionally good value.