With the Oxford Francis Bacon yet to be completed, after more than 150 years the standard scholarly edition of Francis Bacon’s complete works is still The Works of Francis Bacon, which appeared between 1857 and 1859 in seven tomes. ‘A monument to Victorian scholarship’, the Works was the result of a collaboration between three Trinity men – James Spedding (1808–1881), Douglas Denon Heath (1811–1897) and Ellis.Footnote 1 Each was responsible for his own part of the edition – Ellis for the philosophical writings (I-III, 1857) and translations (IV-V, 1858), Spedding for the literary writings (VI-VII, 1858–59) and Heath for the professional legal works (VII). The Works nonetheless came to be associated mainly with Spedding, and this with good reason: among other things, it was Spedding who saw all volumes through the press, who defended the edition against criticism, who completed the edition with his seven-volume The Letters and the Life of Francis Bacon (1861–74), and who gave his blessing to the more popular American editions and abridged reprints.Footnote 2 There is some evidence, however, that the original idea for the Works itself had come from Ellis rather than from Spedding. Already in 1847, Ellis, then a young man of about thirty, was preparing an edition of Bacon’s philosophical works, offered for publication to Longman. Spedding who acted as intermediary between Ellis and the London publisher, reported on the episode in a letter from May 1847 to William Hepworth Thompson, later Master of Trinity College (between 1866 and 1886):

[I] have written to Longman, reporting Ellis’s proposition, and recommending them to treat immediately with him upon those terms; for that if they get the philosophical works alone so edited, their edition will command the market, even if they do nothing but reprint the rest as they are. Whereas, if any other publisher should engage Ellis’s services for that portion, their trade edition would be worthless for ever. […] If they refuse the opportunity, I think I shall decline further connexion with the enterprise.Footnote 3

In the autumn of 1847 a plan was drawn up. It was changed several times over the course of the 1850s and the first volume only saw the light an entire decade later, in 1857. This was due primarily to Ellis’s illness, combined with his initial reluctance to leave his work imperfect or abandon it to someone else even when he was housebound in Trumpington, near Cambridge. Ellis, whose philosophical part was to come first, had already advanced so far in 1847 that he expected to have it ready for the press in 1848. But around the end of 1849, with his Trinity fellowship having just expired and his “Bacon” still unfinished, he was seized with a violent rheumatic fever in Italy, where he had travelled partly on account of his health and partly to do library research.Footnote 4 The last sentence that he wrote would be printed on page 100 of the first volume of the Works, appearing in the preface to the Novum organum. ‘Again he affirms that he does not inculcate, as some might suppose, a –’, to which Spedding attached a note, saying: ‘Mr Ellis had written thus far when the fever seized him’.Footnote 5 It was only in 1853, when he was no longer able to write, that Ellis handed all his papers over to Spedding, with permission to do with them whatever he thought best. Spedding, who in 1842 had resigned his post at the Colonial Office to devote himself entirely to Bacon, decided to take up Ellis’s part himself, ‘not knowing of any one [else] who was likely to take so much interest or able to spend so much time in the matter’.Footnote 6 Ellis did not abandon his work lightly. The proposal to which he agreed in 1853 was that Spedding would ‘print his [Ellis’s] notes and prefaces exactly as I [Spedding] found them; explaining the circumstances which had prevented him from completing or revising them, but making no alteration whatever (unless of errors obviously accidental […]) without his express sanction’.Footnote 7

The outcome of the agreement was that the first five volumes of the Works were not so much co-edited by Ellis and Spedding as they were Ellis’s edition of Bacon’s philosophical work as edited by Spedding. Firstly, Spedding corrected errors, added bibliographical information and penned more than half of the prefaces. He even came up with the tripartite division of the philosophical works: writings which were either published or intended for publication as parts of the Great Instauration (Part I, Volume I-II), independent texts connected with it but not meant to be included in it (Part II, Volume III) and writings originally designed for it but superseded or abandoned (Part III, Volume III). Secondly, and arguably more significantly, in his many notes, prefaces and appendices, Spedding also took the liberty to object where he had anything to object to Ellis’s editorial commentary, though always making clear what was and what was not his. Hence, the Works not merely presented what was then believed to be the complete Bacon corpus. There also was an editorial dialogue running through it, in which Ellis and Spedding debated their interpretations of the nature and relevance of Bacon’s natural philosophy, notably his scientific method and natural histories. This editorial dialogue is fascinating, particularly because Ellis’s and Spedding’s positions were deeply at odds with each other. The difference itself is largely explained by the fact that both had originally arrived at Bacon ‘by entirely different roads’ and studied his work ‘in pursuit of different objects’.Footnote 8 Whereas Spedding sought to vindicate Bacon’s name – which he essentially did by driving a wedge between Lord Verulam’s life and work –, Ellis aimed to ‘penetrate the secret’ of Bacon’s philosophy.Footnote 9 Both men entered the mid-nineteenth-century debate on Bacon by reading Thomas Babington Macaulay’s 1837 review of Basil Montagu’s Life of Lord Bacon.Footnote 10 Spedding thereafter made it his life’s task to defend Bacon against Macaulay, who had praised Bacon’s intellect as a philosopher but ridiculed his moral character as a statesman. He set out his arguments in 1847 in the two-volume Evenings with a Reviewer, which he printed privately and circulated among friends, including Ellis, who lent his copy to Whewell.Footnote 11 It must thus have come as quite a shock and disappointment to Spedding to find out that his co-editor Ellis more or less agreed with Macaulay’s conclusion about the historical significance of the Baconian inductive method: namely, that it had played no role whatsoever in the development of modern science since the time of that other illustrious Trinity man, Isaac Newton.

These, then, were the stakes of Ellis’s and Spedding’s editorial dialogue in the Works: both put forward a view of Bacon’s philosophy – with Macaulay’s views in the background – that they believed to be closest to the spirit and the letter of Bacon’s writings. Like the towering figure of William Whewell (1794–1866), Ellis placed the inductive scientific method – found primarily in the Novum organum – at the heart of Bacon’s entire project. His main argument about it was twofold. First, Macaulay was wrong about Bacon: he had not reduced induction to unaided common sense. Macaulay, in failing to recognize the distinction between enumeration (i.e. observing instances of x) and exclusion (i.e. finding out which laws govern x) as inductive procedures, had merely reinforced the popular misconception that Baconian induction entirely shunned theory and method. Second, Macaulay was right to have pointed out that, taken at face value, Bacon’s method was historically irrelevant and scientifically useless. Rather than abandoning it, Ellis took it upon himself to insist that and suggest how Baconian induction should be renovated if it were to be relevant and useful. (Whewell’s crowning volume of 1858 is tellingly titled Novum Organon Renovatum, with obvious reference to Bacon’s work.) On the basis of a sweeping combination of textual evidence and interpretation, Ellis sought to make an end to Bacon’s ‘Protestant democracy of the intellect’, where anyone with only ‘ordinary acuteness’ and ‘patient diligence’ was able to make scientific discoveries simply by following the inductive method.Footnote 12 It was to be replaced by the aristocracy of genius, where those with the ‘subtlety’ and ‘inventive aptitude’ of a Kepler, a Newton, and a Brewster reigned supreme. Taken together, as Whewell wrote approvingly in 1858, no maxims will ever ‘elevate a man of ordinary endowments to the level of a man of genius’.Footnote 13 There were two parts to Ellis’s renovation, as put forward in the Works: one negative or destructive and the other positive or constructive. First, the features central to the ‘mechanical mode of procedure’, that is, the logic or art, of Bacon’s method – the focus on enumeration and exclusion and the doctrine of forms or simple natures – were to be abandoned or at least toned down.Footnote 14 Together, these had given the false impression that discoveries could result from the mere systematization of facts, rather than requiring what Whewell in his 1857 review of the Works called ‘invention–mind–genius’.Footnote 15 Second, the creative process of the formation of new scientific conceptions had to be given pride of place. Bacon himself had recognized the importance of conceptions but had not written anything on the way in which the inductive method was to be used in their formation. Ellis therefore felt confident to conclude that ‘Bacon never, even in idea, completed the method which he proposed’.Footnote 16 Hence, a renovation of the Novum organum was not merely philosophically and historically needed – as Whewell had already pointed out at numerous occasions –, careful reading – something that Macaulay had failed to do – showed that it could be carried out with Bacon’s own sanction.

Spedding would praise Ellis’s work as an editor, writing in a letter from December 1863 to Harvey Goodwin – Ellis’s “official” biographer – that he had been ‘a good deal disappointed to find the nature, the value and the amount of [Ellis’s] labours upon Bacon so imperfectly appreciated by the popular critics, and so much less preeminence given to his name than was due to it’.Footnote 17 But Spedding was not at all convinced by Ellis’s “Whewellian” interpretation of Bacon’s work, especially because it was at odds with Bacon’s intellectual activities during the last few years of his life, roughly between 1620 and 1626. Most importantly, Ellis had neglected the fact that Bacon had abandoned the Novum organum, instead deciding to complete De Augmentis Scientiarum and start writing Sylva sylvarum, among many other works of experimental natural history. According to Spedding, this shift of emphasis demonstrated that, at least to Bacon himself, natural history was, and always had been, more central to the ‘Great Instauration’ than the inductive method on which Ellis, and Whewell with him, put the sole emphasis.

Despite Spedding’s scruples, during the mid-nineteenth and twentieth centuries, discussions of Bacon and Baconianism would almost exclusively be concerned with scientific methodology.Footnote 18 Ellis’s views, running through the first three volumes of the Works, did much to cement this long-standing orientation. They were focused almost exclusively on the Novum organum and largely neglected, not to say downgraded, Bacon’s natural histories. It is only since fairly recently that Bacon studies has exorcised the ‘ghosts of nineteenth-century interpretation’, contextualizing Bacon’s project within early modern philosophy rather than taking him as an almost contemporary figure.Footnote 19 One outcome is that, after more than one and a half century, the position of Spedding in the editorial dialogue with Ellis is being rediscovered, albeit critically.Footnote 20

1 Origins: Whewell and the Trinity Circle

Rather than fleshing out the details of the editorial dialogue between Ellis and Spedding, the aim in what follows is to trace some of the steps whereby Ellis arrived at what will be called his ‘Baconian idealism’.Footnote 21 The period covered is roughly that between 1833, when the 15-year-old Ellis read Bacon by himself for the first time, and 1853, the year in which Ellis, then still only in his thirties, handed over his papers on Bacon to Spedding. During the interim period, Ellis wrote a student paper that staged a dialogue between Bacon and Newton, discussed the historical scientific relevance of Bacon’s inductive method with Whewell and published two articles on probability theory containing some clues to his own philosophical outlook. The centrepiece of Ellis’s philosophy as well as his Bacon scholarship is without a doubt the ‘General preface’ to the philosophical volumes of the Works, for which he is still remembered today. About a century after its appearance, for example, it was praised by Georg Henrik von Wright, Wittgenstein’s successor as Professor of Philosophy at Cambridge, as ‘the best account of Bacon’s contributions to the logic of induction’.Footnote 22 The 50-page preface is, in fact, Ellis’s only text on philosophy, published or unpublished, and the single available document of his mature philosophical thought. Because its motivations and sources are relatively obscure today – and since explicit references to contemporary authors in the text itself are very scarce – it is not easy to grasp what his mature thought exactly was. It was influenced by Whewell’s work, particularly his History and Philosophy of the Inductive Sciences (published respectively in 1837 and 1840), and contained strong Whewellian elements. At one point, Ellis asked for Whewell’s approval on a matter of interpretation; and Whewell praised Ellis at various occasions, writing in his On the Philosophy of Discovery that ‘Mr. Ellis has given a more precise view than any of his predecessors had done of the nature of Bacon’s induction and of his philosophy of discovery’.Footnote 23 But all this is only part of the story, for there is also evidence, for instance, of Ellis’s possible influence on Whewell and of Whewell’s disagreement with Ellis on specific points regarding Bacon.Footnote 24 Above all, Whewell’s philosophical work was shaped by influences – such as Plato, Bacon, Coleridge and Hare, that is, by Trinity – which shaped Ellis’s thinking too, before he read or met Whewell and even before he went up to Cambridge.Footnote 25

Perhaps the central feature of Ellis’s outlook is that it was nurtured in the large family home in Bath, reading Plato with his classical tutors, and behind the walls of Trinity College, where an anglicized version of idealism held sway since the 1810s–20s. Some called it ‘Platonico-Wordsworthian-Coleridgean-anti-Utilitarianism’; others preferred ‘Germano-Coleridgianism’.Footnote 26 When seen through a Whewellian lens, one might also call it ‘Baconian idealism’. The religious-philosophical outlook of the ‘Trinity Circle’ was an idiosyncratic mix of Platonism, German literature (Goethe, Schiller), history (Schlegel, Niebuhr) and metaphysics (particularly Kant), Romantic poetry (William Wordsworth, Coleridge), and strong Anglicanism. Its key members were Whewell, Julius Charles Hare (1795–1855), Hugh James Rose (1795–1838) and Connop Thirlwall (1797–1875). There were inside outsiders, like the older Trinity College Fellow Adam Sedgwick (1785–1873), and outside insiders, such as Richard Jones (1790–1855), who was at Gonville and Caius College, as well as Trinity apostates who drifted off in a completely different direction – Charles Babbage (1791–1871) being a prominent case in point. Among its younger, second- and third-generation, members were F.D. Maurice (1805–1872), John Grote (1813–1866), J.B. Lightfoot (1828–1889) and Ellis. There were differences and disagreements between all these men. Therefore, both the explanatory power of the label (‘Trinity Circle”) and the extent of its influence (‘Cambridge Network’) should be approached with caution.Footnote 27 But for someone like Ellis at least there was enough intellectual sympathy to share in a sense of common cause. The whole Trinity Circle was united in a quest to defend the eighteenth-century Anglican status quo – with the National Church and National Constitution securing a deeply hierarchical social order – against the connected threats of materialist science, utilitarian morality and democratic politics. Their main aim was to revitalize what was traditional, making the quest for change redundant, but also to guard against what was new, undermining its potential influence. This was done by spreading an “authentic religious temperament” in all domains of human thought and activity, from education to theology and from language to natural science. At least one central feature of authenticity was the unity of poetry and science, of artistic creation and scientific knowledge: both were viewed as asking for a mental activity that is essentially dialectic and synthetic, developing through a process that brings together the personal and impersonal, subjective and objective, spiritual and physical, emotional and logical.Footnote 28

From his days as an undergraduate at Trinity College to his final years as the institution’s Master, Whewell regarded the circle’s quest in terms of the reform of the inductive philosophy, a reform meant to provide the tools and to lay the groundwork for the overarching attempt of giving ‘a right and wholesome turn to men’s minds’.Footnote 29 Whewell’s desire, which he shared with Jones, and initially also with John F.W. Herschel (1792–1871), mirrored – and was influenced by – Francis Bacon’s project, itself – as Hare reminded Whewell in a letter from June 1837 – ‘the fruit of the College, and not of some rubbishy house in Hyde Park Street’.Footnote 30 For them, induction was the ‘true faith’.Footnote 31 Whewell, for one, contrasted the moral and religious power of the personal process of inductive reasoning with the goal-oriented selfishness and atheism of deductive methods, whether in Ricardian political economy or Continental pure mathematics.Footnote 32 However, since Bacon’s own inductive philosophy was flawed, before the Baconian gospel could be preached the “true idea of induction” had to be settled. From around 1830, but starting as early as in 1817, Whewell, reflecting on the (progress of the increasingly) mathematical sciences neglected by Bacon, moved towards his idealist Baconianism: he began to see, and in the rest of his large oeuvre would defend, the idea of induction as an antithetical process involving both an empirical and a conceptual element, the one supplied by careful observation and the other by the creative human mind. Herschel would object that Whewell had, hereby, de facto abandoned Bacon. Some, like George Peacock, held that Whewell was ‘bedeviled with German philosophy’.Footnote 33 Whewell himself, however, firmly believed that his position was fully consistent with Bacon’s own work and remained ‘most true, philosophical, and inductive’.Footnote 34 Like Ellis would do in his ‘General preface’ to the Works, Whewell claimed that Bacon had recognized but not given enough ‘weight or attention to the ideal element in our knowledge’.Footnote 35 Having toyed with the idea for several years, by 1837, when the young Ellis was in his second year, Whewell formulated his position more systematically, appending ‘Remarks on the Logic of Induction’ to his textbook Mechanical Euclid and publishing his History of the Inductive Sciences in that same year. Whewell situated himself somewhere between Locke and his followers of the sensationalist school, who focused only on the empirical, and Kant and the German idealists, who focused only on the ideal. What he believed set him apart was the notion that knowledge requires a combination of both elements: ‘without our ideas, our sensations could have no connexion; without external impressions, our ideas would have no reality’.Footnote 36

Whewell, however, was not a British Kantian of sorts. Neither was he a ‘Germano-Coleridgian’. He read Kant, in a time when German philosophy was not really popular in England, and he appreciated the Critique of Pure Reason, if only for showing that the Circle’s arch-enemies, Locke and his French advocates, were wrong about knowledge. And although he never came as close to Coleridge as, say, his friend Julius Hare, Whewell did acknowledge that Coleridge’s influence on Cambridge philosophy – and that of Trinity in particular – ‘has been so great, and in many respects so beneficial’.Footnote 37 More interesting, also with an eye to his intellectual ties with Ellis, is where Whewell deviated from Kant. Like Kant, he wished to understand how universal and necessary scientific knowledge (e.g. in arithmetic, geometry, mechanics, physics) is at all possible. And his answer, much like Kant’s, involves certain a priori conceptions or ideas that function as conditions of knowledge. One of the major differences with Kant is that Whewell did not distinguish between the a priori elements of knowledge provided by human intuition (Sinnlichkeit), understanding (Verstand) and reason (Vernunft), instead calling space, time and causality ideas, all of which he placed in the human mind. Another difference is that Whewell believed that ideas did not function as conditions of experience per se, but as conditions for knowledge within a particular science (e.g. the idea of Chemical Affinity in chemistry). Consequently, ideas emerged over time and new ones could arise with scientific progress. Perhaps most importantly, Whewell denied that we can only have knowledge of the world as mediated by our own ideas; these ideas represent the world as it is, independent of, and external to, our mind. Whewell’s argument for this view relates to his non-Kantian take on the divine nature of human ideas:

The human mind can and does put forth, out of its natural stores, duly unfolded, certain Ideas as the bases of scientific truths; These Ideas are universally and constantly verified in the universe; And the reason of this is, that they agree with the Ideas of the Divine Mind according to which the universe is constituted and sustained.Footnote 38

This was the Baconian idealism that Whewell placed at the heart of his contribution to the Trinity Circle’s mission and for which Ellis would always feel deep sympathy. All genuine knowledge is inductive but it is the human mind that plays the key role in the process of induction. On the one hand, pace Bacon, the mind is thus an active and creative participant in the attempt to gain knowledge of the world. On the other hand, pace Kant, the ideas through which the human mind obtains knowledge of the world are part of that same world, since both are divine creations. Hence, induction incorporates religion into the heart of science: it is through this gradual process that we ‘learn something of God’ (e.g. that ‘the Ideas of Space and Time are Ideas according to which God has established and upholds the universe’) and that we become aware of ‘the infinitely limited nature of the human mind, when compared with the Creative or Constitutive Divine Mind’.Footnote 39

Ellis’s philosophical outlook, as it developed between the 1830s and 1850s, showed a striking and growing affinity with the views en vogue at Trinity from the 1810s–20s onwards. As a boy, Ellis was trained to study at Cambridge; but – a born Platonist with a German tutor (“Mr Reichel”) – it was almost as if he was made for Trinity. At any rate, it is hard to avoid the impression that there was something about Ellis, whether it were his inborn interests or natural sympathies, that made Trinity the perfect place for him, and vice versa. For instance, it is striking that, before going up there in 1836, Ellis was already well-versed in Plato, Bacon, Berkeley, and Coleridge and that, as a Fellow, he never studied closely J.S. Mill’s System of Logic, instead deciding to read Aquinas, Leibniz, Kant and Dugald Stewart.Footnote 40 Harvey Goodwin, writing in his ‘Biographical memoir’ of Ellis, noted that ‘there was something congenial to [Ellis’s] own cast of mind in the discussions of the schoolmen [the medieval Scholastics, who attempted to reconcile ancient philosophy with Christian theology, LV], and […] he appreciated Bacon all the more in virtue of his appreciation of those, whose processes of thought and methods of argument it was Bacon’s task to supersede’.Footnote 41 There was, indeed, something old-fashioned about Ellis’s bent of mind. But it was precisely this that made him feel at home at Trinity in the 1830s–40s. All of Ellis’s Baconian idealist viewpoints bore the stamp of the Trinity Circle and may be found, in more or less explicit form, in the oeuvre of Whewell, who became his brother-in-law in 1858. Ellis also made original contributions of his own. In fact, he may be seen as the last major advocate of the Trinity Circle before his more prominent friend John Grote (1813–1866), Whewell’s successor as the Knightbridge Professor of Moral Philosophy in 1855, who himself always professed to have been greatly influenced by Ellis.Footnote 42 Ellis would defend its religious-philosophical outlook in other and new contexts beyond poetry, literature and philology (Coleridge, Hare) and history and philosophy of science (Whewell), such as that of pure mathematics and probability theory. Most significantly, of course, it was Ellis who placed it at the centre of what would become the standard Victorian edition of Bacon’s collected works.

2 Philosophy at Bath, 1834–1836

Ellis’s home education as a boy was focused on the Cambridge curriculum of classics and mathematics. Although his wide reading included drama, novels, poetry, ‘useful knowledge’ and sermons, it was Latin, Greek, algebra and geometry that took up most of his time. The very first diary entry, of May 1827, when Ellis was 9 years old, mentions Richard Whately’s (1787–1863) Elements of Logic (1826), to which he would return in July 1832. This work almost single-handedly put traditional syllogistic (or ‘Aristotelian’) logic back on the intellectual map in Britain, where it had fallen into disrepute ever since the second half of the seventeenth century, largely due to the criticism of Bacon and Locke. Bacon had presented his Novum organum as a replacement for Aristotle’s Organon, which contained the six standard books on syllogism. According to Bacon, the formal logic of the Scholastic-Aristotelian tradition was useless for natural philosophy, mainly because it only concerned itself with the validity of conclusions and not with the establishment of their premises. What was needed was an inductive method, applicable to the new sciences, aimed at the discovery of new truths about the world. Whately took it upon himself to defend deductive reasoning against the attack of the British empiricists, even going so far as to deny that induction was a form of inference at all. His radical position did not attract any followers, but almost every nineteenth-century logician would attribute the revival of logic to Whately, and much of British logic from that period may be understood as a commentary on the Elements of Logic – ranging from William Hamilton’s quantified formal logic to George Boole’s algebraic logic and Mill’s and Whewell’s new inductive canons. Since the book itself was largely a criticism of the Baconian tradition, those who defended induction against Whately tended to place themselves in this tradition, whether they were empiricist-utilitarian, like Mill, or idealist and Anglican, like Whewell. At Trinity College in the 1820s–30s, Whately, Nassau Senior (1790–1864) and like-minded ‘deductive savages’ working in logic and political economy from Oriel College, Oxford, were being chastised as ‘ungodly’ and ‘downward mad people’.Footnote 43

Whewell and Jones made it their personal mission to defend the ‘great cause’ – that of a reformed Baconian-inductive philosophy – against deductivist manifestos like the Elements of Logic.Footnote 44 It is rather fitting that this was the first philosophy book that Ellis read, partly by himself and partly with his father. Interestingly, the second philosophy book mentioned in Ellis’s diary was also a work on logic: Isaac Watts’ Logick (first published in 1724, with twenty later editions), which Ellis, following Jeremy Bentham, called the ‘Old-Woman’s Logic’ (17 June 1830). This standard textbook, found by Ellis at the Bath Royal Literary and Scientific Institution, closely followed the syllogistic tradition on reasoning and argumentation, while also sprinkling insights from the work of Locke – for instance agreeing that all ideas come from sensation and reflection. Ellis would read Locke’s An Essay Concerning Human Understanding (1689) together with his father in the summer and fall of 1834, remarking in his diary that he liked a lot of it but that he was undecided about the first book, ‘Of Innate Notions’ (10 August 1834). At several universities, including Cambridge, Watts’ Logick would soon be replaced by Whately’s Elements of Logic, which criticized Bacon, Locke and Watts for trying to introduce the formation of premises into the province of logic. Instead, Whately argued, induction is an inference governed by syllogistic form, such that the conclusion is always already contained in the premises. Like Whewell, Ellis would come to disagree with both approaches, as they failed to capture the imaginative mental process by which knowledge is obtained. If this did not have a place in mainstream logic, whether inductive or deductive, then, as Whewell would remark in an 1859 letter to his former pupil Augustus De Morgan, ‘so much the worse for Logic’.Footnote 45

Before reading Bacon, or Whewell, Ellis read “Herschell’s” Preliminary Discourse on the Study of Natural Philosophy (1831), which had the bust of Bacon on its title-page. (Ellis started reading on 4 January 1831 and finished the 350-page book a week later.) Like Whewell, Herschel wished to renovate Bacon’s inductive scientific method, but they had different views on what this method entailed.Footnote 46 These different views of Bacon are apparent in the reviews they wrote of each other’s books. Whewell, writing in the Quarterly Review, accused Herschel of neglecting Bacon’s ‘condemnation of the method of anticipation, as opposed to that of gradual induction’ in allowing for hasty leaps to hypotheses. Some nine years later, Herschel returned the charge of anti-Baconianism in his lengthy joint review of Whewell’s Philosophy and History of the Inductive Sciences, lamenting the ideal or conceptual element of Whewell’s position. As he wrote to his friend: ‘You are too a priori rather for me – as soon as one has worked one’s way up to a general law you come cranking in and tell me it is a Fundamental Idea innate in everybody’s mind’.Footnote 47 Ellis enjoyed the Preliminary Discourse when he first read it. Some twenty years later, however, he confessed to James D. Forbes that he thought Herschel was ‘a charlatan: honorably distinguished no doubt […] but neither clear nor deep’.Footnote 48 It is interesting to observe that in the Works, Spedding would refer to Herschel as part of his argument against Ellis, who, tellingly, always only referred to Whewell when the opinion of a ‘modern scientific writer’ was needed.Footnote 49

More than Aristotle, Herschel, logic or method, the young Ellis liked Plato and metaphysics. ‘Every man is born an Aristotelian, or a Platonist’, wrote Coleridge in 1830 – and, as Hare told Whewell in 1822, ‘Plato is worth ten thousand Aristotles’.Footnote 50 There is some truth to this as far as Ellis is concerned. He would eventually read Aristotle – initially his dramatic and literary theory and later his philosophical works, in the context of editing Bacon – but never expressed a favorable opinion of his work.Footnote 51 By contrast, from early on, Ellis was duly convinced that knowledge was not limited to observable things; there were abstract truths too, which could be apprehended by some higher faculty than reasoning. Some of these may well be innate (10 August 1834). At the same time, there were things, like the real nature of the soul, of which he knew that they were ‘above human intelligence’ (27 March 1833). Due to his awareness of their mutual influence, Ellis’s major philosophical concern in his 1830s diaries was the Platonic theme of the relation of body and mind, of the physical and the ideal, to one another. ‘I am not in spirits, another black fit; surely there is a connection between the mind & the body’, Ellis, for instance, noted on 12 October 1834: ‘If the mind be an essence, sui generis, it is odd that it grows with the growth of the body – that it waxes and wanes as the physical energies’ (12 October 1834). Together with Cicero (whom he first read on 1 January 1831), particularly his De finibus bonorum et malorum (begun in August 1833), where the views of Epicureanism, Stoicism and later Platonism were explained, Plato was the young Ellis’s first and main philosophical influence. This is not surprising per se, as Ellis, of course, came to philosophy through his classical tutor Johnston. After several years of Homer, Ovid, Virgil, Aeschylus and the likes, in November–December 1831 Ellis, aged 14, construed his first “Plato”, the Second Alcibiades, a dialogue on prayer.Footnote 52 Plato immediately became a favourite. A year later, in December 1832, Ellis noted: ‘I should very much like to read some more of Plato. The dialogue on to kalon (‘the beautiful’) [Hippias Major, LV] I should like to understand. The phrase “love of the beautiful” seems to me tautology’ (14 December 1832). (Ellis read the whole of Hippias Major in September 1834, by which time he criticizes its anatreptic form.) By January 1833 he began the Phaedo with Johnston, the famous dialogue on the immortality of the soul, something which he tellingly took be ‘rather natural than shown by reasoning’ (7 January 1833). Ellis liked both the style and ‘sentiments’ of it, writing that the ‘metaphysical reasoning’ was even better than in Cicero’s De natura deorum (On the Nature of the Gods) (16 and 21 January 1833). ‘What it is that makes me so fond of metaphysics I know not’, Ellis wrote in his diary in January 1833; what he did know was that he admired Plato ‘fervently’ and that he understood the arguments better than Johnston (25 and 29 January 1833, 15 February 1833). (About a year later, he concluded: ‘I am not fond of any but the most abstract science’, preferring that which has ‘the least reference to any actual existence’ (9 October 1834.)) By May 1833, Ellis wrote confidently that he found some of Plato’s reasoning ‘very loose’ – all the same, ‘I would if I could know what the author is about’ (3 and 17 May 1833). Around that time, Ellis was already reading other, early modern philosophy works – apparently at his own initiative and of his own choosing. For example, on 5 May 1832 he noted that George Berkeley’s Three Dialogues (1713) – a semi-popular defense of the idealist epistemology advocated in the canonical Treatise Concerning the Principles of Human Knowledge – ‘seems to have been taken more from Cicero than from Plato’ (5 May 1832). Ellis returned to Berkeley 1 month later, reading the Alciphron, or The Minute Philosopher, a plea against Catholic ‘freethinkers’. ‘I do not attempt to understand [it] at all thoroughly, though I admire the subtlety of what I can comprehend’, Ellis noted on 16 June.

Bacon entered the scene in Ellis’s diary in the Spring of 1833, when Ellis, aged 16, began a translation into Greek of Bacon’s essay ‘Of Empire’ (1612). A few months later, in October, he started to work on a re-translation into Latin of a passage from (what must have been a later English translation of) the ‘beautiful’ ‘Wisdom of the Ancients’ (De sapientia scientiarum, 1609) (23 October 1833). Since he came to Bacon, and Plato, through Latin and Greek writing exercises for Johnston, Ellis was much concerned with Bacon’s style, again using Cicero as a standard of excellence. For example, on 3 November 1833 he notes: ‘I wish that I cld write as this great man whose style reminds me more of the gorgeousness of Cicero than any author I remember’ (3 November 1833). A week later he notes: ‘We are so flippant, so missish [i.e. prim or characteristic of a young lady or schoolgirl] in our writing these days that the golden style of Bacon & Milton & Taylor is neglected & even censured’ (9 November 1833). Throughout his diaries, Ellis also constantly reflects on the content of Bacon’s works, writing what he likes about, say, the ‘Wisdom of the Ancients’ (‘there is great ingenuity in these interpreted myths’, 19 May 1834) and asking himself why he likes the Essays even better (‘the influence of these short & pithy sentences, into which much thoughtful observance of humanity is compressed, can hardly be estimated’, 20 May 1834). For Ellis, style was always intrinsically related to content, in so far as beauty and knowledge were, for him – as they had been for Plato and Coleridge before him – more or less inseparable. At times, Ellis even wrote of the ‘worshipping of the beautiful’ (11 May 1834) as a condition for arriving at true knowledge. He ascribed an epistemic status to poetry, as Coleridge had done in his Lyrical Ballads (1798), or at least shunned the distinction between the poetic and scientific. Despite all his merits, Bacon was not a poet: he lacked both the ‘taste’ and ‘exquisite sense’ to recognize the ‘connection & sympathy between the great world & the microcosm of the heart’ (11 May 1834). This is interestingly related to another, more personal, remark from later that year, where Ellis reflected on Bacon’s personality. He called Bacon ‘a man who was all mind’, one who ‘preferred the intellectual pleasures to those of the affections’ (22 June 1834). It was not that Ellis himself did not prefer the contemplative life, but he simply experienced too much physical and mental discomfort to be entirely ‘bookish’. In September of that year, for instance, he remarks that pain ‘is absolutely necessary to our well-being; that restlessness is part of our essence’ (16 September 1834). The underlying message was that what someone knows is not only reflected in that person’s style but also conditioned on how he or she has come to know it – a personal process to which the ‘connection between the mind & the body, which conveys the impressions made on the one to the other’ is crucial (11 October 1834). All the same, whether someone arrives at knowledge of outward things has everything to do with that person’s inner life, as ‘moral worth’ and ‘intellectual superiority’ are inseparable too (27 July 1834). ‘I am fully convinced’, wrote Ellis, ‘that a thoroughly good intellect is always accompanied by pure & holy morality’, as ‘a heart corrupt argues a head unsound’ (27 July 1834). Ellis here touched on what would become two major themes of the early Victorian era: the notion of genius and its connection with method and morality, such as in the case of Newton, and the tension between Bacon’s moral character and philosophical work.Footnote 53

It was in May 1834 that Ellis for the first time read one of Bacon’s philosophical works systematically and all by himself: ‘The Advancement of Learning’ (1603) – a work later expanded and Latinized as De augmentis scientiarum (1623). This formed the first part of the ‘Great Instauration’, containing a division and classification of the new sciences organized around the tree metaphor: the sciences are like branches of a tree that meet in a stem, a single “universal science’, or philosophia prima, whose axioms are shared by all the individual branches. ‘No one can admire Bacon more than I do – his felicity of expression & his wonderful perspicacity are enough to deserve all the admiration of posterity’, wrote Ellis on 12 May 1834. ‘Nevertheless’, adding a remarkably prescient remark:

I think that the “Principia” […] would have been written had the highpriest of philosophy never existed or if circumstances had prevented him writing the “Advancement” or the “Novum Organum” (12 May 1834).

By that time, Ellis was sporadically reading ‘ye “Organum Novum”, which he finished in August of that year (see 29 June and 3 August 1834). This unfinished work – the second part of the Great Instauration, originally published in 1620 – contained Bacon’s inductive scientific method, which (much to Whately’s ridicule) was meant to replace in toto the Aristotelian syllogistic method. Ellis found it rather ‘tiresome’ at times, especially its long tables of heat, whereby Bacon illustrated the process of the induction by exclusion (29 June 1834). What interested Ellis were not the applications and examples but the abstract fundamental principles of science. When it came to these principles, the Novum organum, however ‘noble’, was ‘full of errors – the errors of the time’ (3 August 1834).

Ellis was not only reading Bacon’s own works but also some of the major Victorian commentators on Bacon. Already in February 1834, he ‘looked over’ Basil Montagu’s Life of Francis Bacon, quickly concluding that Montagu, the editor of the then standard edition of Bacon’s writings (16 volumes, 1825–37), ‘does not seem a great scholar’ (11 February 1834). Although he might not have thought much of his intellectual capabilities – let alone of his knowledge of Greek and Latin –, it was through reading Montagu that Ellis entered the nineteenth-century debate on Baconianism and, thereby, eventually arrived at Bacon scholarship. For it was Montagu’s Life that was the object of Macaulay’s criticism and it was Montagu’s complete edition of Bacon’s writings that the Works sought to replace.

‘I think that my studies are over for the present – as I have little left to read’ (28 September 1834). After 1834 Ellis, indeed, read but little philosophy, as his Cambridge preparations intensified and his focus came to lie even more exclusively on classics and mathematics. He occasionally returned to Plato and Bacon, and there were Alexander Pope’s An Essay on Man (1734) (19 June 1836), which popularized the ‘optimistic’ view, found in Newton and Voltaire, of a divinely ordered human condition, and Seneca the Younger’s De Vita Beata, a book on the good life which Ellis did not like (‘It is astonishing how little there is in it. Epicurus was the only one who has caught a glimpse of the truth – that all motives spring from a consciousness of pleasure’, 13 June 1836). One other exception was the Hare brothers’ Guesses at Truth (first anonymous edition in 1827), which Ellis found in a London library in July 1836 (15 July 1836). (Ellis would receive the book as a gift from his friend Tom Taylor, who was enthusiastic about Hare, several years later, in February 1839.) He praised it for its strong ‘contempt for “ratiocinative understanding”, which furnished him with much ‘material for thought’ (15 July 1836). The less philosophy Ellis read, the more he turned to penning down his own thoughts on philosophical matters, albeit sporadically: in August 1834, he resolves to write ‘a metaphysical essay to show my profundity’ (1 August 1834) and in May 1836 he is scribbling remarks on the Novum organum (8 May 1836).

Ellis’s main philosophical influences in the 1830s came from books, as well as from Johnston and his father, with whom he read some of them. One other personal influence from this period deserves mention, that of the Rev. H.F.C. (Henry) Logan (1800–1884) – Professor of Mathematics (1830–?) in the Catholic College of Prior Park, near Bath, and President of St Mary’s College, Oscott (1847–8). Logan, a correspondent of William Rowan Hamilton and Arthur Cayley, among others, and an acquaintance of Davies, Ellis’s mathematics tutor, was well-versed in Continental mathematics and German philosophy. By 1834, having reflected on ‘the theory of Idealism’, found in the work of Berkeley, Kant and Fichte, he came to hold the belief ‘that the phenomena of the external world and their laws are the produce of the mind itself’, adding that ‘the Theory of Space and Time adopted by Kant […], which I believe to be the only sound one yet proposed, evidently leads to the same conclusion’.Footnote 54 Ellis was introduced to Logan in the summer of 1834, when for some time the two met almost weekly, and he much enjoyed their conversation. Logan amused him with a curious story about the Polish mathematician-philosopher-messianist Jozéf Maria Hoene-Wronski (1776–1853) and his “Absolute philosophy”. It was also Logan who first told Ellis about Kant’s Critique of Pure Reason. In a diary entry of 24 June 1840, Ellis would note that ‘I caught the germ of it, viz. Kant’s view of time & space, from Logan ages ago’ (24 June 1840). He then went on to remark that he had often ‘sported’ (i.e. proposed, tried out) the idea that ‘time cannot be a cause – because time is only the condition of the existence of causes’ in examinations, wondering whether ‘I made my fortune with Whewell by so casual an advantage’ (24 June 1840). Among the many things that Ellis brought with him from Bath to Cambridge, Kantian sympathies evidently were one of them.

3 Philosophy at Cambridge, 1836–1840

At Trinity College, Cambridge, where Ellis came into residence in October 1836, the focus was on classics, in the first year, and on mathematics, from the second year on. One of the rare occasions on which Ellis returned to philosophy – or to Bacon, more particularly – as a student was for the English declamation competition (the ‘Hooper Prize’) of 1838. These prizes, endowed in 1760 by Francis Hooper (1694–1763), were to be awarded each year to three students for ‘an English Oration to be delivered publickly in the [College] chapel on a Subject entirely relative to the English Nation or History’.Footnote 55 As a phenomenon in itself, the Hooper Prize is quite remarkable for several reasons. One is its ‘Englishing’ of declamation at Trinity College, where Latin was the traditional standard, going back to the statutes of 1560. Another, more striking, reason is its proof of the continuing importance of rhetorical activity in the University at a time when mathematics and written examinations came to dominate at the expense of Cambridge’s oral academic culture. For a student like Ellis, it may be added, the competition provided a welcome escape from mathematics (of which he was ‘sick to death (23 December 1838)) and an opportunity to develop his own thoughts, some of which had ripened for several years.

Ellis began to prepare for his declamation in November 1838, writing out the first half on the 9th – rehearsed the next day (‘just about remembered it’) –, finishing a complete version on the 24th – read to his friend Alexander Gooden (1817–1841) on the 27th (who said that ‘it has too much matter, & is too difficult’) –, and having a fair handwritten copy made to submit to the Dean, before finally delivering it in the College Chapel on 1 December.Footnote 56 Ellis, who would rank second after Lord Francis Napier (1819–1898), agreed with Gooden that his text did not make for a good declamation, considering that it were the ‘flimsy things which get prizes’.Footnote 57 But, writing to his father, Ellis decided that ‘I would rather write what would show that I understand & have thought upon a difficult & important subject’ (i.e. ‘the authority of induction’), to which he added that ‘if [William] Carus [(1801–1891), Senior Dean of Trinity College], who is the judge, understands my declamation, he is more of a philosopher than I take him for’.Footnote 58

Ellis’s 1838 declamation on Baconian induction seems to have been inspired by two texts: Macaulay’s 1837 review of Montagu’s Life of Francis Bacon and Coleridge’s 1818 ‘General Introduction, or, Preliminary Treatise on Method’. Macaulay had influentially attacked both Bacon’s moral character and his philosophical reputation. About this famous review, Ellis would at one point make the telling and prescient remark that, ‘as for Bacon’s philosophy, he does not understand it’s [sic] spirit’ (6 April 1840). Coleridge, for his part, had made a rather striking attempt to dissociate Bacon from his French admirers, notably Condorcet and other followers of Locke, by presenting Bacon as the “British Plato”. Both great philosophers had been misunderstood: Plato never denied the importance of experience and, despite what the French philosophes said, Bacon had not rejected the contribution of ideas, of ‘an intellectual or mental initiative’ in arriving at scientific knowledge.Footnote 59 (It was this two-sided, or “antithetical”, method that Coleridge proposed as the plan for the entire Encyclopaedia Metropolitana project, which was to replace the French Enlightenment encyclopedias, as well as the Encyclopedia Britannica and Chamber’s Cyclopedia.)Footnote 60 ‘Of Bacon’s character as a man […] it is not my intention to speak’, the 19-year old Ellis opened his text: ‘It is as the author of the Novum Organum that I shall attempt to consider him’.Footnote 61 What followed were some twenty pages where Ellis took up a question he had already raised in his diary in 1834. It is typical for the nineteenth-century focus on Baconian scientific methodology that Ellis would himself help put in place as editor of the Works: is it possible to ascribe the progress of modern science to the inductive method? Does the Novum organum justify the claims set up on its behalf? Or, that is: ‘Is the method of philosophizing advocated by Bacon the same as that which has led to our brilliant advancement in the natural sciences?’. The question is approached through an imaginary dialogue between Newton and Bacon. Rather than an open conversation, however, the dialogue essentially provides a list of reasons why Bacon’s idea of scientific method cannot be reconciled with, and is even in conflict with, Newton’s scientific practice (despite the famous hypotheses non fingo promise):

He would assuredly deprecate the use of explanatory hypothesis; he would reject [the] guess that gravity may extend to the moon as an “anticipatio mentis”; […] he would deplore the absence of any system of “prima philosophia”, the want of method, the neglect of tabulation; […] above all, he would complain that we reason too much, and observe too little.

Unlike others – Whately in the 1820s, Macaulay in the 1830s, Augustus De Morgan in the 1850s, etc. – Ellis did not point to Newton to abandon Bacon. Instead, he drew lessons from Newtonian practice to modernize Baconian theory. Since Bacon could not be blamed for the source of his mistakes – namely, ‘the want of a real knowledge of the history of any part of science’ – this strategy seemed justified.Footnote 62 On the one hand, Ellis agreed with Bacon’s attack on deductive syllogistic reasoning, writing that ‘the subtlety of nature is far greater than the subtlety of logic’ (7). It is impossible, even for Newton, to discover new truths about the world by following laws or rules describing how certain conclusions are to be drawn from given premises. In fact, ‘truth is sui generis subject to no laws’ (7) and the ‘historical element’ of knowledge cannot be neglected.Footnote 63 On the other hand, what Newton’s practice also showed is that it is not by the systematization of observation that scientific discoveries are made. Ellis, echoing the Coleridgean language of Guesses at Truth – in a manner strikingly similar to Julius Hare’s pupil Maurice’s The Kingdom of Christ, also from 1838 –, believed that the history of science showed that this is done by strokes of brilliance, flashes of light and ‘sparks of celestial fire’ (11):

[M]an’s mind does not allow him to see into the mysteries of nature except in those fitful glimpses which occasionally visit the higher order of intellects and with a conviction of truth which they cannot communicate to others, nor justify to themselves. Still the light has shined, and by a deductive process we may verify what men call a simple hypothesis, by comparing the consequences which would flow from it with results of observation, and thus the intuition of genius is converted into the common property of mankind. (10-11)

What does it mean to incorporate the mental aspect of Newtonian practice into the Baconian inductive method? Ellis gave one example, touching on a topic that he would return to in his 1842 essay ‘On the foundations of the theory of probabilities’ (read before the Cambridge Philosophical Society in February and published in its transactions in 1844): the ground of induction, which is neither inductively nor deductively established but, as Ellis argued in opposition to the Lockean school of Condillac, a priori, relating to ‘the constitution of the human mind’ (13).Footnote 64 ‘The authority of induction is founded on a conviction that the antecedent conditions being alike in two cases, the subsequent state will be so too’ (12). This conviction is not the outcome of observations of a certain event or of the calculation of the degree of belief that it will happen again on the next instance. It results from the human mind’s contemplation of ‘all the cases under consideration’, eventually recognizing – or, as Ellis writes, ‘intuiting’ – their ‘entire similarity’ (12–13). Ellis’s position at this point echoed the Hares’ notion of an ‘idea’ or ‘divination of the whole’ which allows an individual thinker to penetrate a given object much deeper than through the ordering and arranging any of its specific features.Footnote 65 Ellis’s 1842 paper gave another example where this is brought out even more clearly. One notion central to the classical probability of Jacob Bernoulli, Pierre-Simon Laplace and others was the idea that an event will, on a long run of trials, tend to recur with frequency proportional to its probability. ‘This is generally proved mathematically [by means of the law of large numbers]. It seems to me to be true a priori’:

When on a single trial we expect one event rather than another, we necessarily believe that on a series of similar trials the former event will occur more frequently than the latter. The connection between these two things seem to me to be an ultimate fact […], the evidence of which must rest upon an appeal to consciousness. Let any one endeavor to frame a case in which he may expect one event on a single trial, and yet believe that on a series of trials another will occur more frequently.Footnote 66

Ellis called such beliefs ‘axioms’ – in a letter to Ellis, De Morgan spoke of “latent axioms”: ‘things which at first are not even credible, but which settle down into first principles’ – and held them to be a priori truths, ‘supplied by the mind itself”.Footnote 67 As Ellis would remark in a diary entry from August 1840, ‘it is the characteristic of necessary truths to become intuitive to a mind familiarized to them’ (26 August 1840).

Bacon had put too much emphasis on observation and too little on concepts; Plato vice versa. But both had recognized that knowledge was two-sided, as it always involved both the factual and the mental. Hence, in this crude sense, there was a connection to be made between Bacon and Plato (and Kant), at least according to Coleridge, Maurice and like-minded Trinity men.Footnote 68 Ellis knew of this connection; he referred to the rather idiosyncratic view of Bacon as the “British Plato” in his 1838 text, albeit critically. About a year later, however, he showed much more enthusiasm, noting in his diary on June 1839 that ‘from what Coleridge says of the possibility of reconciling Plato & Bacon, I conjecture that he had a sounder view of Bacon’s philosophy than is usual’ (28 June 1839). (Ellis would return to this interpretation in the Works, where he, like Coleridge, wrote that abstract forms, rather than concrete phenomena, were the object of Bacon’s philosophy.) Another month later, in July 1839, Ellis read Henry Hallam’s (1777–1859) account of Bacon in his Introduction to the Literature of Europe (1837). Here, Hallam – the father of Arthur Henry Hallam, friend of James Spedding, Lord Tennyson and their Trinity set – spoke very highly of Bacon’s philosophical writings, both of its style and content. Ellis appreciated Hallam’s ‘high recommendation of Bacon’s Essays’, adding that Bacon’s mind was, indeed, ‘all vigour’ (25 July 1839).

Like so many of his contemporaries, by the late 1830s, Ellis was in two minds about Bacon: the Novum Organum was without a doubt ‘a great work’ (2), but also hopelessly flawed; and Bacon himself, ‘the most brilliant writer on Philosophy’, was ‘rather admired than understood’ (20). Ellis would draw what he saw as the only logical conclusion: Bacon’s flaws should be remedied by a proper understanding of his oeuvre.

4 Ellis’s Philosophy in the 1840s–50s

A few months after he graduated Senior Wrangler, in January 1840, Ellis sat the Trinity Fellowship examination, which included several questions on philosophy, among other topics; two on Plato, one on Aristotle and two more on Cicero.Footnote 69 His preparation for the philosophical part seems to have involved little more than studying some of Plato’s dialogues, including the Gorgias and the Theaetetus, and reading Friedrich Schleiermacher’s (1768–1834) Introductions to Plato’s Dialogues (‘I did not make much of it’, 20 June 1840) and Johann Gottfried Stallbaum’s (1793–1861) prolegomena attached to his edition of Plato.Footnote 70 But the very decision to want to obtain the Fellowship, to which he was duly elected on a first attempt in October of that year, was important, as it shows that Ellis had decided to make Cambridge his home. During the interim period, which he knew was ‘a critical moment’ (28 February 1840), he was without ‘schemes of future life’ (25 August 1840), living from day to day: he read some mathematics and philosophy, mostly ethics – often ‘unable to fix on any subject, or follow out a single train of ideas’ (28 February 1840) –, worked on ‘some trifles’ (28 March 1840) for the Cambridge Mathematical Journal (CMJ) and very occasionally scribbled notes on metaphysics (e.g. 29 March and 5 June 1840), none of which have survived.

The only project for which Ellis seems to have been able to enthuse himself in these months was that of turning the reflections on probability theory from his 1838 declamation into a paper for the CMJ.Footnote 71 It was the writing of what would become his 1842 paper ‘On the foundations of the theory of probabilities’ that led Ellis into his first intellectual exchange with Whewell, which would gradually turn into a collaboration. Ellis proposed the idea of an essay on probability to Duncan Farquharson Gregory (1813–1844) – one of the founders of the CMJ alongside William Walton (1813–1901) – on 3 April 1840. A few days later, he received a letter from Walton, recommending him to wait for Whewell’s Philosophy of the Inductive Sciences to appear before publishing on the subject. Ellis himself had already thought of writing to Whewell, which he did on 7 April, ‘stating my notion of writing a little essay on probabilities, & asking if his work on the Philosophy of induction would interfere with it’ (9 April 1840). Two days later, Ellis received ‘a very civil answer’: Whewell ‘was very glad to hear of my intention & wished me to persevere as “he was sure I would throw light on it”’ (9 April 1840). As Ellis wrote in his diary: ‘Whewell’s letter showed that one of the most arrogant of men of science is ready to acknowledge me as a fellow labourer in a favourite field of speculation’ (9 April 1840). Ellis immediately set to work, writing on probability and reading Whewell’s Philosophy, about which he soon concluded that ‘very many of the views [expressed], are so far my own, that they have, more or less perfectly formed, passed through my own mind’ (14 June 1840). A first version of the paper was completed in January 1842, when Ellis shared it with Whewell, saying that he was prepared to revise or change anything that Whewell saw fit.Footnote 72 Whewell, who himself wrote next to nothing on probability, approved of it and was instrumental in its delivery to the Cambridge Philosophical Society, where Ellis read it on 14 February. Whewell’s approval was hardly surprising, as the paper was essentially an attempt to re-establish the foundations of probability theory on a Whewellian basis, that is, to reconcile it with ‘a philosophy of science which recognizes ideal elements of knowledge, and which makes the process of induction depend on them’.Footnote 73 Ellis was by now a wholehearted public supporter of this philosophical position. A few months earlier, upon reading Berkeley’s De Motu (1721) – where Newton’s absolute space, time and motion are called into question –, he remarked: ‘[The book] is of little value, except as showing how enormous is the common cry of our natural philosophy being founded entirely on experience. A subtle enquirer into what experience really is, will either recognise the existence of principles in our knowledge not derived from experience, or will be led like Berkeley into destructive scepticism’ (20 May 1841). It was in this same period that Ellis wondered whether his Kantian view on space and time might have influenced how Whewell wrote about the first law of motion in his Philosophy.Footnote 74

There is no decisive evidence on when Ellis started working on Bacon. By 1847, he expected to have his part of the Works ready for the press within a year or so. The papers which he would hand over to Spedding in 1853 contained the ‘General preface’ (50 pages), some ten prefaces (about 90 pages in total), numerous notes and a selection of translations which together would ‘give an English reader a complete view of the Baconian philosophy’.Footnote 75 Given the kind of expertise required – historical, bibliographical, scientific, philosophical – and the vast amount of literature needed to be read – from Bacon to Aristotle, from Leibniz to Vossius, from Plato to Paracelsus and from Sextus Empiricus to Thomas Aquinas – it is likely that Ellis started somewhere in the mid–1840s at the latest. By that time, his attitude towards Bacon must have been even more ambiguous than before. He had now embarked on a project dedicated to a philosopher whom he revered but whose oeuvre he found deeply flawed; in 1849, he travelled to the Continent to work on the Novum organum, a book he had denounced as being ‘full of bad philosophy & impractical rules’ (10 March 1841). And in the early 1850s, when he was already very ill, he was still eager to give at least some attention to Bacon. During the last years of his life, Ellis ‘could not bear the subject of Bacon to be alluded to: if it happened to be introduced, he would say, “We don’t talk about it in this room”.Footnote 76 There evidently was something to Bacon that captivated him, that made him forgive him all his errors, and that gave him a motive to re-edit the Baconian corpus. Ellis himself offered a clue in his ‘General preface’ when he wrote that Bacon’s ‘merits ‘belong to the spirit rather than to the positive precepts of his philosophy’.Footnote 77 This spirit was a religious vision of the aim of natural philosophy and of the unity of science and religion, of knowledge and faith:

[Bacon] did good service when he declared, with all the weight of his authority and of his eloquence, that the true end of knowledge is the glory of the Creator and the relief of man’s estate. When, not long before Bacon’s time, philosophy freed itself from the tutelage of a dogmatic theology, it became a grave question how their respective claims to authority might be most fitly coordinated. It was to meet, perhaps rather to evade, this question, that the distinction between that which is true in philosophy and that which is true in religion was proposed and adopted. But it is difficult to believe that the mind of any sincere and truth-loving man was satisfied by this distinction. Bacon has emphatically condemned it. ‘There is’, he affirms, ‘no such opposition between God’s word and His works’. Both come from Him who is the Father of lights, the fountain of all truth, the author of all good; and both are therefore to be studied with diligence and humility.

There is some truth in the idea that this was also what Ellis, and like-minded Fellows of Trinity with him, ‘wished earnestly to say […] to the men of his generation’.Footnote 78 What was distinctive about Ellis’s religious-philosophical outlook was not its agreement with anything Bacon said, however; it was the Baconian spirit in which it disagreed with, and went over and beyond, Bacon. Ellis, like Whewell, introduced ideas as the basis of all inductive knowledge and, by that fact alone, found himself in a different position than Bacon vis-à-vis the unity of science and religion. Their non-Baconian task was to understand how it was possible that, in the progress of science, it turns out that human ideas correspond to the facts, and the facts to human ideas. And their non-Baconian answer was that, since both are of divine origin, they cannot but be in total agreement. Or, as Ellis, aged 38, formulated it in one of his last papers, of November 1854:

Man in relation to the universe is not spectator ab extra, but in some sort a part of that which he contemplates […]. The thoughts we think are, it is true, ours, but so far as they are not mere error and confusion, so far as they have anything of truth and soundness, they are something and much more. […] In every science the fact and the idea correspond because the former is the realization of the latter, but as this realization is of necessity partial and incomplete […] this correspondence is but imperfect and approximate.Footnote 79