A Cartesian challenge to the early modern philosophy of experiment

Much has been written about seventeenth-century experiments and experimental philosophy. My paper for the CELFIS seminar of October 8 aimed at engaging with that tradition. In particular, I was concerned with the recent discussion by Peter Anstey of the so called BBH model of the experimental philosophy (BBH stands for the name of Bacon, Boyle, and Hooke). As a reaction to Thomas Kuhn and Peter Dear, Peter Anstey’s article provides a very nice introduction into the Baconian experimentation and its main developments in the second half of the seventeenth century. Both Boyle and Hooke engage with a form of experimentation that is labelled here “Baconian.” It is not, however, the purpose of this small blog post to engage with the details of Anstey’s article, but rather to try to complement his analysis with a new example of experimentalism that can be found in a completely different source. This is the case of the experimental work of the Cartesian natural philosopher, Jacques Rohault.

In my lecture, I’ve referred to two experiments that were performed by Rohault: with pneumatic devices, on the one hand, and with glass drops, on the other hand. It is well known that Boyle was the main contributor to the pneumatic or baroscopic experiments of the 1660s. Hooke was among the first to examine glass drops and to provide an explanation for both the production of the small glass objects and for the curious phenomena produced by those. Interestingly, Rohault deals with both of these issues in experimental terms.

Now, one might very well wonder why is important that a Cartesian philosopher was providing an explanation for some intriguing experiments; after all, he is a Cartesian, therefore a speculative philosopher (see the Otago blog here and here), and he would explain all phenomena according to the principles of Cartesian physics. Yet, this classification of seventeenth-century philosophers into “experimental” and “speculative” should not be an impediment in searching for explanations in one’s writings. But there is more than that and I argued in my paper that it is precisely Rohault’s experimental approach to the study of the two phenomena that would make difficult to draw a clear boundary between his work and the works of the most representative experimenters of the BBH model.

I have argued elsewhere that Rohault treats the study of the properties of the air in experimental terms. He does not simply jump from the conclusions derived in the general part of Cartesian physics (which is most often claimed that he does), but actively engage in experiments and observations.

With respect to the study of glass drops, Rohault is also concerned to perform all the needed observations before providing an explanation. This is also what Hooke did in his Micrographia.

As a tentative conclusion for this very sketchy blog-post, I claim that based on these two experiments, Rohault should be placed in the same context with Boyle and Hooke, so as a representative of the BBH model. If, on the contrary, one would like to point to his “Cartesianism,” then, one simply overlooks his experiments and this would raise new worries for the use of historical categories: if one dismisses some experimental practices only on the basis of placing the practitioner to one or another camp, then, the problem is not any more with the use of experiment in natural philosophy, but with the way various natural philosophies of the period were classified in our histories.

Francis Bacon and the use of measurement in experiments

One central component of experimental philosophy is measurement. Various properties, quantities, degrees and qualities were counted and measured with more or less exactitude starting with the early modern period (see for example, the analysis of temperature measurement in A.Borrelli, “The weatherglass and its observers in the early seventeenth century”, in: Claus Zittel, Gisela Engel, Nicole C. Karafyllis and Romano Nanni (eds.), Philosophies of technology: Francis Bacon and its contemporaries, vol. 1 (Leiden: Brill, 2008) 67-130 (Intersections 11/1)). Seen in itself, measurement was almost universally considered a tool meant to improve knowledge and to give strength to different arguments and to rebut others. This attitude was shared by Francis Bacon too, as I briefly attempted to show in a small presentation I have made for the 4th Bucharest Colloquium in Early Modern Science. One can infer this from the following example taken from Bacon’s Sylva Sylvarum:

“It is strange how the ancients took up experiments upon credit, and yet did build great matters upon them. The observation of some of the best of them, delivered confidently, is, that a vessel filled with ashes will receive the like quantity of water that it would have done if it had been empty. But this is utterly untrue; for the water will not go in by a fifth part. And I suppose that that fifth part is the difference of the lying close or open of the ashes…” (Sylva Sylvarum, SEH 34).

There are many other examples of experiments in which Bacon used to invoke the measurement and counting of quantities in order to champion his ideas (see for instance, entries 1, 19, 21, 32, 33, 46, 59, 60, 76, 88, 104-110, 156, 159, 248, 306, 307, 309, 310, 318, 324, 363 etc, to give just few examples taken from the first three centuries of Sylva). Here are two more extended examples:

“Dig a pit upon the sea-shore, somewhat above the high-water mark, and sink it as deep as the low-water mark; and as the tide cometh in, it will fill with water, fresh and potable… I remember to have read that trial hath been made of salt water passed through earth, through ten vessels one within another, and yet it hath not lost his saltness, as to become potable: but the same man saith, that (…) salt water drained through twenty vessels hath become fresh… But it is worth the note, how poor the imitations of nature are in common course of experiments, except they be led by great judgment, and some good light of axioms. For first, there is no small difference between a passage of water through twenty small vessels, and through such a distance as between the low-water and high-water mark…” (Sylva Sylvarum, SEH 1-2)


“The continuance of flame, according unto the diversity of the body inflamed, and other circumstances, is worthy the inquiry; chiefly, for that though flame be (almost) of a momentary lasting, yet it receiveth the more and the less: we will first therefore speak (at large) of bodies inflamed wholly and immediately, without any wick to help the inflammation. A spoonful of spirit of wine, a little heated, was taken, and it burnt as long as came to one hundred and sixteen pulses. The same quantity of spirit of wine mixed with the sixth part of a spoonful of nitre, burnt but to the space of ninety-four pulses. Mixed with the like quantity of bay-salt, eighty-three pulses. Mixed with the like quantity of gunpowder, which dissolved into black water, one hundred and ten pulses. A cube or pellet of yellow wax was taken, as much as half the spirit of wine, and set in the midst, and it burnt only to the space of eighty-seven pulses. Mixed with the sixth part of a spoonful of milk, it burnt to the space of one hundred pulses. Mixed with the sixth part of a spoonful of water, it burnt to the space of eighty-six pulses… So that the spirit of wine simple endured the longest; and the spirit of wine with the bay-salt, and the equal quantity of water, were the shortest.” (Sylva Sylvarum, SEH 366)

I propose the following table as a tool for a concise representation of Bacon’s measurement of the continuance of flame:

  • Spirit of wine = 116 pulses
  • Spirit of wine + 1/6 nitre = 94 pulses
  • Spirit of wine + 1/6 bay-salt = 83 pulses
  • Spirit of wine + 1/6 gunpowder = 110 pulses
  • Cube of yellow wax + ½ spirit of wine = 87 pulses
  •  Cube of yellow wax + wine + 1/6milk = 100 pulses
  • Cube of yellow wax + wine + 1/6 water = 86 pulses


This raises a plenty of interesting questions dealing with the way Bacon uses measurement that is worth to be discussed. Here is a tentative list with few of them:

–          What type of measurements does Bacon employ?

–          What type of quantities or qualities is subjected to measurement by Bacon?

–          What other examples of Bacon’s measurements can be represented by such tables?

–          In what theoretical cases is measurement invoked?

–          How important is the exactitude in the measurements?

–          When does measurement help in constructing an argument and when does it help in rejecting other’s arguments?

–          Is measurement an effective tool in building up the theory of matter?

–          Is measurement used independently of Bacon’s theory of matter?

The examples of measurements Sylva Sylvarum presents can be a good starting point for the discussion of these points. Different answers to these questions can also set the stage for a comparative analysis between Bacon’s use of measurement and other philosophical treatments of it.

We would love to hear your comments, suggestions and thoughts on these matters, so please leave us a comment.

IV. Seventeenth-Century Experiments with Glass Drops: Robert Hooke on glass drops

In the previous posts, I have introduced the problem of glass drops in the seventeenth-century natural philosophy, which I have further discussed in the case of Jacques Rohault’s Cartesian experimentalism and other Cartesian explanations provided by Henricus Regius and Nicolas-Joseph Poisson.

In the final post on this topic, I would like to refer to the explanation provided by an experimental philosopher par excellence, Robert Hooke (1635-1703).

Some of the earliest investigations of the drops took place in the Royal Society (see Brodsley, Laurel, Charles Frank, and John Steeds. 1986. “Prince Rupert’s Drops.” Notes and Records of the Royal Society of London 41 (1) (October): 1–26). Hooke was one of its most distinguished members and witnessed these inquiries from the very beginning (see Birch, Thomas. 1756. History of the Royal Society. Vol. I. London: A. Millar). After an initial report signed by Robert Moray in 1661, Hooke published his own observations in the celebrated Micrographia of 1665. He applies jointly direct observation, hypothetical thinking, and experiment in order to formulate an explanation “Of some Phenomena of Glass drops,” as it is said in the title of his seventh observation. As expected, glass drops and fragments of drops are examined through the microscope. Yet, Hooke notes the difficulties of his inquiry: “I could not find, either with my naked Eye, or a Microscope, that any of the broken pieces were of a regular figure, nor any one like another, but for the most part those that flaw’d off in large pieces were prettily branched” (p. 33). Trying to find the imperfection of glass, sets Hooke on a quest of successive trials, such as grinding the object in different parts and observe if there are new effects or immersing glass drops in various substances (e.g., Icthyocolla). Like in the prior cases of Rohault and Regius, Hooke complements his text with an illustration (see Fig. X at the bottom of this image):

Robert Hooke, Micrographia (1665).

The object is not only investigated empirically with various instruments, but the ‘conjectures’ formed in this process are further examined through analogy and transduction, which ports theoretical elements into Hooke’s analysis. Moreover, descending into the invisible structures of matter, forces him to leave the eye and the microscope and frame his explanation on top of other prior considerations about heat, fluidity, and matter in general. He concludes by ascribing the cause to the arrangement of particles of glass, which are gathered into a springy tension inside the drop.

It is not the place in this blog-post to go into the details of Hooke’s explanation of the phenomena produced by glass drops, but it is worth noticing that his experimental methodology overlaps with what I have presented in the post dedicated to Rohault. Theory and experimentation work together in finding the explanation, which, ultimately reduces to a mechanical model of the accepted theory of matter. These four cases of seventeenth-century natural philosophers interested in the study of glass drops make wonderful examples of how “new scientific objects” are discussed and explored with new methodological tools. I shall investigate this problem into a forthcoming article.

II. Seventeenth-Century Experiments with Glass Drops: Jacques Rohault and his Cartesian experimentalism

In the previous post, I’ve only introduced glass drops as objects of philosophical study in the early modern Europe. I argued that they have become popular in the late 1650s and early 1660s, at the same time with the spread of a more extensive experimental attitude in the investigation of nature. In this post, I shall give a brief account of Jacques Rohault’s empirical investigation of these objects.

Rohault (1618-1672) was one of the leading Cartesian philosophers of his time. He hosted a famous salon in Paris, where he jointly discussed problems of natural philosophy in connection with experiments. Moreover, his book on physics – the Traité de physique (1671) – was quickly adopted as a textbook on natural philosophy in various universities. As I argue in a forthcoming article, much of Rohault’s experimental work has been done in the late 1650s and early 1660s (see Dobre, Mihnea. “Rohault’s Cartesian Physics.” In Cartesian Empiricisms, eds. Dobre, Mihnea and Nyden, Tammy. Springer Studies in History and Philosophy of Science. Springer). Among his experiments, Rohault performed some with “larmes de verre” (see Christiaan Huygens’s report from March 17, 1660: “Rohaut [sic], qui fit l’experience d’une larme de verre” in Huygens, Oeuvres XXII, p. 562).


Jacques Rohault. Traité de physique (1671), p. 173.

For Rohault, glass drops (“larmes de verre”) are wonders of nature, which means that they challenge natural philosophy. Stating the extraordinary character of the object from the very beginning allows him to test the explanatory power of his own natural philosophy. This comes with both a theoretical and experimental dimension. First, Rohault draws the boundaries of his explanation in terms of what is theoretically acceptable; in other words, a general description of how motion works. Then, Rohault performs several observations, showing: (1) the quick cooling of glass; (2) that glass-surface becomes cooler, which he already explains by the fact that it closes its pores; (3) matter inside is still agitated producing both glass powder and larger pores; and finally, (4) the question: why is not breaking?

All sorts of experimental trials accompany the first three steps of this investigation. Rohault applies various operations to the larme de verre (e.g., attempts to break the drop in other points than D or to immerse the object in different substances and observe changes in the properties of the glass), which can be considered variations in the experimental procedure. He concludes his empirical investigations by answering the fourth point from above. Glass drops are difficult to break on point D, because of the arrangement of matter: the inner part of the drop is formed by a mixture of air and glass – with some sort of arch-structures – while the external part is very dense. This is easily explained by the process of fabrication: when melted glass is left to plunge into a flask of cold water, it quickly cools, such that glass particles that are in contact with water will solidify immediately, while the inner parts will continue their motion.

Using jointly the principles of Cartesian natural philosophy and an experimental method, Rohault is forced at this point to delve into the hidden structure of bodies. He appeals to Cartesian theory of matter (body is identical with matter and space; void space is not possible) and mechanical modeling of interacting bodies. Yet, this is done in a similar manner with his contemporaries, the so-called experimental philosophers.

In the next post, I shall discuss the cases of Henricus Regius and Nicolas Poisson. That will conclude my overview of Cartesian natural philosophers dealing with the problem of glass drops.

I. Seventeenth-Century Experiments with Glass Drops: an introduction

One of the new ‘scientific’ objects that captured the attention of the seventeenth-century French and British experimentalists is what they call a ‘glass drop’: the result of dropping a bit of incandescent glass into a bucket of cold water. Peculiar to this object was its solidity, which was contrasted with the very minute fragmentation of glass that was produced when the drop was finally destroyed.

A representation of the glass drop can be observed in the following illustration from Jacques Rohault’s Traité de physique (1671).

Jacques Rohault. Traité de physique (1671), p. 173.

This is the first post of a series discussing some seventeenth-century attempts to explain the properties of glass drops. Considered as one of the most intriguing objects, glass drop became an interesting item of study for many natural philosophers, including Robert Boyle, Jacques Rohault, Henricus Regius or Robert Hooke. As its name suggests, the object was made of glass and its peculiar shape intrigued as much as its apparently contradictory properties. On the one hand, the glass drop was found very difficult to break when it was pressed on the large size of the drop. On the other hand, its tail was very easy to fracture, which produced the blow of the drop with a loud noise.

A modern replica of the phenomena – starting from the production process of the drop and continuing with an exemplification of the two properties described above – is possible to watch online on the website of the Corning Museum of Glass (see http://www.cmog.org/video/prince-ruperts-drop): http://youtu.be/6V2eCFsDkK0.

The glass drop became philosophically interesting only in the seventeenth century. References to its structure and properties appeared almost at the same time in several places of Europe. The drop had also different names: larme de verre, lacryma Batavica, chymical glass, Prince Rupert’s drop, vitrae lacrymae, globuli vitrei, Batavian tears, etc. Its origins are still controversial, but it is commonly held that it came from either German lands or the Dutch provinces. I shall not try in these blog-posts to discuss the origins of this object or to attempt to correct some misconceptions about who has the priority in the philosophical investigation of the drop (an overview of these problems is in Brodsley, Laurel, Charles Frank, and John Steeds. 1986. “Prince Rupert’s Drops.” Notes and Records of the Royal Society of London 41 (1) (October): 1–26.). Rather, by looking at some of the early discussions and explanations of the phenomena produced by glass drops, I would like to suggest that early modern authors were forced to tackle the problem in a manner that included both theoretical and experimental examinations. This makes the glass drop case important for any discussion about the sources of knowledge in early modern philosophy, complicating the traditional divide between Rationalists and Empiricists and shading a new light on other more recent attempts to use actor-category terms, such as experimental and speculative (see the research project on Early Modern Experimental Philosophy at the University of Otago: https://blogs.otago.ac.nz/emxphi/the-project/).

In the next post, I shall discuss Jacques Rohault’s explanation of the phenomena.

Sylva sylvarum: experiments on transmutation of bodies (24.04.2012)

This meeting followed two important directions: 1) the relation between Bacon’s matter theory and Sylva Sylvarum Century I and 2) the particular discussion of experiments 25 to 30 and the way they connect to the other experiments of Century I. This post deals with the second one.

(25) Experiment solitary touching the making of artificial springs. Before touching the actual experiment, Bacon makes a series of observations about experiments in general that are worth being mentioned. First, we are told that although it may be unexpected, Bacon actually continually rejects experiments. Yet, he tells us that “if an experiment be probable in work and of great use, I receive it, but deliver it as doubtful”: a) How should we understand Bacon’s skeptical attitude towards experiments? b) Does he have a criterion (or criteria) for a trustworthy experiment? Bacon’s constant skeptical attitude towards experiment shows that he does not accept experiments at face value. He seems to believe that some experiments do not provide certain knowledge, and that one has to be constantly aware of the experiment’s limitations, and of its construction, and how this tool should be used for the study of nature. A good demonstration of Bacon’s constant preoccupations with the limits of experiments is the first set of experiments from Sylva, where he criticizes Della Porta’s experiment of filtration of seawater as a bad case of translation (from a natural fact to an artificial fact). Thus, Della Porta’s experiment is taken to be an unreliable experiment whose results are not trustworthy. Consequently, for Bacon, what is important is not whether indeed sand can filter the seawater and make it potable, but the fact that Della Porta’s experiment is not actually telling us anything about that particular phenomenon.

The experiment of an artificial spring goes as follows: On sloping land, a hole is dug, and in the hole a trough of stone is introduced. The hole is covered with brakes and sand. What it is observed, according to Bacon’s source (Bacon did not actually perform the experiment), is that even after the rain stops, a spring of water can be observed at the lower end of the trough. According to Bacon, if this is the case, then this phenomenon can be read as a case of transmutation of air into water because it is as if “the water did multiply itself upon the air, by the help of the coldness and condensation of the earth, and the consort of the first water.”

If this reading is correct, then this experiment should be correlated to experiment 27 which discusses the version and transmutation of air into water. This experiment is interesting in many respects. The experiment cites 4 processes through which air is transmuted to water, or in terms of matter theory, a more pneumatic body (the air) into a more tangible body (the water). Those 4 processes are: condensation (as the example of experiment 25), compression (e.g. distillation vapors, dews), the mingling of moist vapours with air (method suggested for testing via an experiment), and via the porosity of bodies. A few things need to be noted. Bacon grades those 4 processes differently: the first 2 are apparent and sure, the last 2 are considered probable, but not yet manifested. Bacon deals with study cases of these processes later on in Sylva, in an entry entitled Experiments in consort touching the version and transmutation of air into water, where the experiments from 76 to 82 deal precisely with how the ‘version’ is acquired via such processes. The immediate question raised is why Bacon chose to (or did he actually choose to?) separate experiment 27 from experiment 76–82. In fact, experiment 29, entitled Experiment solitary touching the condensing of air in such sort as it may put on weight and yield nourishment, seems to fit well with these experiments. This entry touches on also a process of transition of air, as a pneumatic body (the air) to a denser, tangible body. Bacon’s reasoning is the following: usually for sprouting, seeds/plants are buried in the ground and watered. Yet, some things sprout even if they are only left in the air. Bacon proposes to verify whether those things that sprout in the air increase in weight, thus gain some solid mass. If they don’t increase in weight, then the sprouting is just an inner transformation of the body. If they do increase in weight, then Bacon reasons that the only place where this new mass could come from is the air, which in return would mean that the pneumatic air has transformed into a denser, tangible body. Some things to be noted here: the conclusion Bacon reaches here is dependent on his matter theory. Moreover, we could see the experimental proposal that Bacon makes here as a case of a corroborative evidence for the problem of transforming air into a denser body. A second thing: this experiment could also be used to study the problem of whether air can nourish or not, which is nothing other than a case of translation, one of the methods of experientia literata.

Coming back to experiment 27, we observe that this experiment includes some methodological moves worth mentioning: One of the examples given as a case of transmutation of air into water via condensation is the following: “and the experiment of turning water into ice, by snow, nitre, and salt, would be transferred to the turning of air into water”. Provided the 8 rules of EL, this would be a case of production by extension, since as exp. 82 claims, it “is a greater alteration to turn (artificially) air into water, than water into ice”. The same experimental setup is used to study two different problems: whereas the transformation of water into ice is a case of induration of bodies (a process happening in the same body), the transformation of air into water is a case of transmutation (a case of transition from one species to another).

If a connection seemed apparent between these experiments, we couldn’t trace any ways to connect experiments 26 and 28.

Experiment 26 entitled Experiment solitary touching the venomous quality of man’s flesh. This ‘experiment’ establishes a correlation between cannibalism and its malignant effect for human bodies on the basis of a collection of reported instances of cannibalism. In this example, we could say, in modern terms, that instances are corroborated and that a fact (that syphilis/“the disease of Naples” was originally caused by cannibalism) is considered to be probable precisely because of its status as corroborated evidence. On the other hand, we failed to see how this experiment connects to previous ones or how this experiment could be suggested by any of the others that Bacon has presented so far. Even more, we failed to see what theoretical question underpins it. This also happened with experiment 28, entitled Experiment solitary touching the helps towards the beauty and good features of persons, where Bacon discusses how some of man’s features, while growing, can be moulded by pressure.

Proposals for connecting some experiments:

a. experiments 25, 27, 29, and 76 to 82 appear to study a similar problem—whether a pneumatic body can be transformed into a tangible one

b. experiments 17–23 and experiments 76, 77, 79, and 80 deal with  the problem of how pneumatic matter is “trapped” into bodies, e.g. in infusion, in the pores of bodies . We also alluded to a connection of these experiments with the experiments on percolation (in Sylva’s text, experiments 1 to 8. )