In retrospect: A New System of Chemical Philosophy

Journal name:
Date published:
Published online

Philip Ball reflects on the work of John Dalton, father of modern atomic theory.

A New System of Chemical Philosophy

John Dalton R. Bickerstaff: 1808.

Visual metaphors are often essential in science when you can't see what you're studying. The English chemist John Dalton, born 250 years ago, illustrated his atomic theory using wooden spheres (pictured), drilled with holes for pins that enabled them to be linked into clusters. But there are hazards to such mental props. By the 1880s, students were so familiar with the spheres that one (taught by prominent advocate of atomic theory Henry Enfield Roscoe) declared: “Atoms are round bits of wood invented by Mr Dalton.”

Ian Dagnall/Alamy

John Dalton, painted in 1835 by Thomas Philips.

Today, the atoms Dalton proposed in his seminal New System of Chemical Philosophy (1808) are routinely revealed by microscopy and crystallography. They are corralled in electromagnetic traps, pushed around like marbles using scanning probe microscopes, even manufactured and monitored one at a time in superheavy forms using particle accelerators. No one mistakes them for bits of wood.

Neither did Dalton. He articulated the ancient idea that matter is built from fundamental particles in a way that aligned it with the quantitative principles of chemical reaction elucidated in the late eighteenth century. Those macroscopic rules, he said, stemmed from the systematic combination of microscopic bodies: solid, massy and hard, as Isaac Newton had put it in a phrase Dalton was fond of quoting.

Yet in a sense, even by the 1880s, atoms were still not much more than Dalton's model spheres. Because they remained unobserved, several leading scientists refused to accept their reality, among them physicist Ernst Mach and chemist Wilhelm Ostwald. Some considered atoms no more than an heuristic convenience: a crutch that the mind could use to make sense of chemical transformations. That is why, despite Roscoe's misgivings that Dalton's wooden balls might mislead students, the balls had a valuable role. They showed how visualizing an entity can help to cement the concept even while direct evidence is elusive. It is a risky strategy to assert the physical reality of something not yet observed (will dark matter really be particulate?). But without such an image, a theory can seem little more than metaphysics.

Manchester Museum of Science & Industry/SSPL

The spheres that Dalton used to demonstrate atomic theory.

It is traditional to locate Dalton's New System of Chemical Philosophy as a step — perhaps the greatest — in a long road to modern atomic theory that began with the ancient Greek atomists Leucippus and Democritus in the fifth century BC, and ended with the nuclear atoms proposed by Ernest Rutherford and Niels Bohr in the early twentieth century, then quantum theory and scanning probe microscopes. The “philosophy” in Dalton's title signified something closer to a scientific theory than to the abstract reasoning it tends to connote today. Yet his book also represents an important juncture for the philosophy of science. It spoke to whether science should be based on empiricism or explanatory hypothesis — a question that had exercised Newton and Robert Boyle in the seventeenth century. There was nothing new in Dalton's idea of atomistic matter; the question was whether to treat this as a useful conjecture or as a reality. Antoine Lavoisier, whose work on the proportions of chemical combination was crucial to Dalton, had no time for such questions. Lavoisier insisted that meditating on “ultimate particles” was metaphysical — and fruitless.

So how did Dalton, a modest teacher educated in Cumbrian village schools and excluded from Oxford and Cambridge for his Quakerism, take an imaginative leap that eluded distinguished professors? Even if we admit some of the fairy dust of “genius” into an explanation, we shouldn't discount Dalton's wide reading — from Boyle and Newton to Claude Louis Berthollet and Humphry Davy. He also paid careful attention to the quantitative details of experiments by the likes of his friend, Mancunian chemist William Henry, and Lavoisier. Dalton presented his atomistic theory to the Manchester Literary and Philosophical Society, of which he was secretary, between 1803 and 1805. Some of his papers were published in the society's memoirs, but he was urged to present them as a book, as he put it, in “the interests of science, and his own reputation”.

The New System is one of those foundational books that doesn't say what you might think it should. It is mostly not about atoms at all. The first 140 pages or so of Volume 1 dwell on heat and its effects, whereas Volume 2 is a detailed account of inorganic chemical compounds. Dalton's atomic theory is confined to the five-page final chapter of the first volume. Here, he explains that the fixed stoichiometries of chemical reactions — so much of element A combines with so much of B — can be rationalized by supposing that the constituent atoms unite into “compound atoms” of simple ratios, such as 1:1 or 1:2. The point is most famously and eloquently made in a plate that shows sketches of these unions. An “atom” of water comprises one atom each of hydrogen and oxygen; an atom of ammonia is a 1:1 union of hydrogen and nitrogen (Dalton uses Lavoisier's term, “azote”, for nitrogen).

The proportions are wrong — chemist Jöns Jakob Berzelius corrected many in the following two decades. And in 1813, he proposed an alphabetical representation (for example, H2O [sic]) in place of Dalton's pictorial balls. Dalton, with the conservatism common to trailblazers, declared this “horrifying”, saying that the symbols “cloud the beauty and simplicity of the atomic theory”. His displeasure might have contributed to a stroke in 1837.

The New System is not a new theory of chemistry. Among other things, it offers no explanation for why atoms react. Roscoe put his finger on it when he said that the significance of Dalton's theory was his proposal that each type of atom has a unique mass. That made sense of the quantities in which elements were found to combine, and offered the first general and fundamental distinction between one element and the next — what eventually became embodied in the idea of atomic number.

Yet it is the idea of atoms as the indivisible units of matter that stuck in the mind, because readers could see them on the page. Dalton didn't intend his pedagogical diagrams of atomic unions — “compound atoms”, or molecules as we'd now say — to be taken too literally. There's no inkling in his book of molecular shape; the arrangements of atoms in binary, ternary and other unions are purely notional, and when Dalton draws “water particles” packed into the crystalline forms of ice, they too are spheres.

All the same, visual representation of atoms was surely the precondition for the emergence of a concept of molecular structure, with atoms in fixed spatial relationships, in the mid-nineteenth century. Something of this kind would surely have appeared whether or not Dalton had “invented” atoms as wooden balls — but that innovation was more eloquent than its inventor anticipated.


  1. Report this comment #68661

    Majid Ali said:

    Atoms, Wood Spheres, and the Pyramid

    In the context of the nature of matter, two competing stories have endured since antiquity: atomism and holisim. The former is transparently directed to structure and the latter implicitly to function. Democritus (460 BC ? 370 BC), a pre-Socratic Greek philosopher and his mentor, Leucippus, are generally credited with visualizing foundational particles that make up all forms of matter. In India, the Jain, Ajivika and Carvaka schools of atomism are claimed to date back to the 4th century BC. Holism (Greek holos, for whole, entire) offers a philosophic-functional counterpoint of matter, and considers structural, biologic, social, and economic matters as systems. It subordinates discreteness of atomism to its core tenet of the whole.

    Philip Ball re-tells the story of wood spheres with central holes with which John Dalton (1766-1844) created as enduring images of atoms as unique masses held together by forces represented by pins in the central holes. He credits this visual representation of Dalton for creating the precondition for the emergence of a concept of molecular structure with atoms in fixed spatial relationships. He recognizes that the visual of spheres had succeeded beyond what their creator might have imagined. He also speculates that if Dalton, the father of modern atomic chemistry and the author of New System of Chemical Philosophy (1808), had not offered his visual, someone else would have done so.

    Ball informs that many errors in the illustrations of Dalton were subsequently corrected by others as he speaks of the enormous explanatory and didactic power of images. Visuals, of course, have not been the exclusive domain of atomists. Holists have been busy at it as well. Since antiquity, stone carvers and painters have offered visuals of roots of trees to point to the sources of sustenance for the leaves and fruit. Leonardo DaVinci stands tall among them. His visual of the Vitruvian Man with outstretched hands and spread legs is hard to escape in holistic health circles. The DaVinci Man is his simplest illustration of a shifting centre (of a holistic view of the human frame).

    In clinical medicine, visual representations for fostering holism in the science and philosophy of healing have been neglected. Most practitioners of western medicine will struggle to think of any visual representation of holism of value in their clinical work. In the 1980s, this writer, a surgeon (FRCS (Eng)-turned pathologist (Columbia University, New York)-turned integrative clinician, shifted from the study of disease to a study of health. During this work, an image of a pyramid of a trio of trios of human ecosystems slowly evolved for establishing clinical priorities. In 2000, this was presented in Oxygen and Aging (ref. 2). The schema of this pyramid (posted online, ref. 3) comprises a base, an apex, and the area between the two with the constituent ecosystems as follows: (1) the base trio of the bowel, blood, and liver ecosystems (with a sharp focus on the issues of altered microbiota); (2) the middle trio of the adrenals, thyroid, and pancreas ecosystems; and (3) an apical trio of the skin, gonadal hormones, and neurotransmitters. The pyramid is lit by a rising sun symbolizing the philosophic-spiritual context of health and healing. A compendium of published clinical outcome studies were subsequently published in Darwin, Dysox, and Disease, the 10th, 11th, and 12th volumes of the Principles and Practice of Medicine (ref.4-6).

    My interest in academic holism began with this random thought several years ago: evolution could not have designed the structures of the human body merely for the expedience of anatomists. Working as hospital pathologists, my colleagues and I considered cells of a specific organ as unique masses, as Dalton visualized his atoms. Clinicians sending us their biopsies expected us to provide precise des criptions of those unique masses which allowed us to make diagnoses of specific pathologic entities. To cite a specific example, neither we pathologists nor colleagues in gastroenterology ever looked for an answer to the problem of colitis in the workings of the stomach, or solutions to gastritis-esophagitis complex in altered states of colon ecology. The idea of ecologic thinking in clinical medicine never crossed our minds before we began to focus on immunopathology of the gut visualized as a single functional unit (ref. 7-10). That work led to adding clinical integrative medicine to our pathology work with a focus on restoring colon flora in the treatment of the gastritis-GERD complex, which soon validated this facet of the Pyramid model. Similarly, investigation and restoration of insulin homeostasis proved valuable in resolving the intractable clinical problems of the thyroid gland, giving merit to considering the thyroid, adrenals, and pancreas as a trio of clinical significance.

    In closing, the Pyramid model has four strengths: (1) it shifts focus from discreteness of morphologically-defined chronic diseases to a molecular (respiratory-to-fermentative shift) model of the essential connectedness of various organ systems (ref. 10); (2) it provides sound scientific basis for focusing on major homeostatic systems of the body (oxygen signaling and insulin homeostasis being the most important in my clinical work); (3) it offers a scientific rationale for incorporating in patient care indigenous therapies with empirical benefits; and (4) it fosters an integrated, holistic-functional view of health and healing in patient education.


    1. Ball P. A New System of Chemical Philosophy. Nature.2016; 537: 32.
    2. Ali M. Oxygen and Aging. New York, Canary 21 Press. Aging Healthfully Book 2000. Pp 285.
    3. Ali M.
    4. Ali M. Darwin, Oxygen Homeostasis, and Oxystatic Therapies. 3 rd. Edi. (2009) The Principles and Practice of Integrative Medicine. Volume X. (2009) New York. Integrative Medicine Press.
    5. Ali M. Darwin, Dysox, and Disease. 3rd. Edi. 2009. The Principles and Practice of Integrative Medicine. Volume XI. New York. Integrative Medicine Press.
    6. Ali M. Darwin, Dysox, and Integrative Protocols. New York (2009). The Principles and Practice of Integrative Medicine Volume XII: New York. Integrative Medicine Press.
    7. Ali M, Mesa -Tejada R, Fayemi AO, et al. Localization of IgE in tissues by an immunoperoxidase technique. Arch Pathol Lab Med, 103:274-275, 1979.
    8. Ali M. Ramanarayanan MP, Nalebuff DJ, Fadal RG, Willoughby JW: Serum concentrations of allergen-specific IgG antibodies in inhalant allergy: effect of specific immunotherapy. Am J Clin Pathol, 80:290-299, 1983.
    9. Ali M, Ramanarayanan MP: A computerized micro-ELISA assay for allergen-specific IgE antibodies. Am J Clin Pathol, 81:591-601, 1984.
    10. Ali M. Fayemi AO, Nalebuff DJ: Localization of IgE in adenoids and tonsils: an immunoperoxidase study. Arch Otolaryngol, 105:695-697, 1979.

Subscribe to comments

Additional data