A First Sketch of the Nature of Heat (Novum Organum Book 2: 15-25)

by Francis Bacon 11d8th Nov 201923 min readNo comments


This is the eleventh post in the Novum Organum sequence. For context, see the sequence introduction. For the reading guide, see earlier posts in the sequence.

We have used Francis Bacon's Novum Organum in the version presented atwww.earlymoderntexts.com. Translated by and copyright to Jonathan Bennett. Prepared for LessWrong by Ruby.

[[In the previous section, Bacon introduced his "three tables": his careful collection of data and observations that are core to building up his scientific method.

These tables are:

1) A table of presence which lists many examples where phenomena of interest in presence, e.g. many examples of things where we have heat.

2) A table of nearby essence. His example is heat and to discriminate its true heat, Bacon looks for examples of things that resemble those in the table of presence yet are lacking the heat. For the example, the light of the moon (cold) is contrasted with the light of the sun (hot) which is interesting given they are both heavenly bodies.

3) A table of degrees or comparison where are examples are brought where the amount of perceived heat differs in degree between things. This is also useful in discriminating the true underlying cause and and nature of heat.]]

Aphorism Concerning the Interpretation of Nature: Book 2: 15-25

15. The job of these three tables is—in the terminology I have chosen—to present instances to the intellect. After the presentation has been made, induction itself must get to work. After looking at each and every instance we have to find a nature which

  • is always present when the given nature (in our present case: heat) is present,
  • is always absent when the given nature is absent,
  • always increases or decreases with the given nature, and
  • is a special case of a more general nature

(I mentioned this last requirement in 4). If the mind tries to do this •affirmatively from the outset (which it always does when left to itself), the result will be fancies and guesses and ill-defined notions and axioms that have to be adjusted daily. (Unless like the schoolmen* we choose to fight in defence of error; and in that case how well an axiom fares will depend ·not on how much truth it contains but· on the ability and strength of its defender.) It is for God (who designed and gave the forms), and perhaps also for angels and higher intelligences, to have an immediate •affirmative knowledge of forms straight away. This is certainly more than man can do. We have to proceed at first through

[[*Schoolmen: Aristotelian scholars.]]


[Bacon will now be likening scientific procedure to a kind of chemical analysis, in which various components of a complex liquid are distilled off by heat, leaving the residue in which we are interested.]

So we have to subject the nature ·in which we are interested· to a complete dismantling and analysis, not by fire but by the mind, which is a kind of divine fire. The first task of true induction (as regards the discovery of forms) is to reject or exclude natures that

  • are not found in some instance where the given nature is present, or
  • are found in some instance from which the given nature is absent, or
  • are found to increase in some instance when the given nature decreases, or
  • are found to decrease when the given nature increases.

After these rejections and exclusions have been properly made, and all volatile opinions have been boiled off as vapour, there will remain at the bottom of the flask (so to speak) an affirmative form that is solid, true and well defined. It doesn’t take long to say this, but the process of doing it is lengthy and complex. Perhaps I’ll manage not to overlook anything that can help in the task.

17. I have to warn you—and I can’t say this too often!—that

When you see me giving so much importance to forms, do not think I am talking about the ‘forms’ that you have been used to thinking about.

·Treating my forms as your ‘forms’ in the present context would be wrong in two ways·. (1) I’m not talking here about composite forms, the ones in which various simple natures are brought together in the way the universe brings them together—the likes of the forms of lion, eagle, rose, gold, and so on. It will be time to treat of these when we come to hidden processes and hidden microstructures, and the discovery of them in so-called substances or composite natures.

(2) In speaking of ·forms or· simple natures, I’m not talking about abstract forms and ideas which show up unclearly in matter if indeed they show up in it at all. When I speak of ‘forms’ I mean simply the objective real-world laws of pure action* that govern and constitute any simple nature—e.g. heat, light, weight—in every kind of matter and in anything else that is susceptible to them. Thus the ‘form of heat’ or the ‘form of light’ is the same thing as the law of heat or the law of light; and I shan’t ever use abstractions through which I step back from things themselves and their operations.

[*Bacon doesn’t explain actus purus. In each of its other three occurrences he connects it with laws, and his meaning seems to be something like: ‘the laws governing the pure actions of individual things, i.e. the things they do because of their own natures independently of interference from anything else’. If x does A partly because of influence from something else y, then x is not purely •active in respect of A because y’s influence gives A a certain degree of •passivity. From here on, actus purus will be translated by ‘pure action’.]

[In the next sentence, ‘rarity’ is cognate with ‘rare’ in the sense of ‘thin, attenuated, not dense’.] So when I say (for instance) in the investigation of the form of heat

  • ‘reject rarity ·from the list of simple natures that constitute heat·’, or
  • ‘rarity does not belong to the form of heat’,

·I may seem to be talking about an abstract property rarity, but what I am saying can just as well be said without any noun purporting to refer to any such abstraction. For· those statements are tantamount to

  • ‘It is possible for us to make a dense body hot’, or
  • ‘It is possible for us keep or remove heat from a rare body',

·where ‘rarity’ and ‘denseness’ give way to ‘rare’ and ‘dense’·.

You may think that my forms also are somewhat abstract, as they mix and combine things that are very different from one another. ·This complaint might come from your noticing that·

  • the heat of heavenly bodies seems to be very unlike the heat of fire,
  • the relatively durable redness of a rose (say) is very unlike the ·transient shimmering· redness that appears in a rainbow, an opal, or a diamond, and the different kinds of death—by drowning, burning, stabbing, stroke, starvation—are very unalike;

yet they share the nature of heat, redness and death respectively. If you do have that thought, this shows that your mind is captive to •habit, to •things taken as a whole ·and not subject to analysis or bit-by-bit examination·, and to •men’s opinions. For it is quite certain that these things, however unalike they may be, agree in the form or law that governs heat, redness and death (respectively); and human power can’t possibly be freed from the common course of nature, and expanded and raised to new powers and new ways of operating, except by discovering of forms of this kind. This •union of nature is the most important thing I have to talk about; but when I have finished with it I shall take up, in the proper place, the •divisions and veins of nature, both the ordinary ·superficial· ones and also the ones that are more internal and true. ·By the ‘union of nature’ I mean the coming together of disparate things under a single form. By the ‘division and veins of nature’ I mean the complexities in which disparate structures and functions come together in a single thing·.

18. I should now provide an example of the exclusion or rejection of natures that are shown by the Tables of Presentation not to belong to the form of heat. All that is needed for the rejection of any nature ·from the form we are investigating· is a single ·contrary· instance from one of the tables; for what I have said makes it obvious that any conjecture ·of the type ‘Nature N belongs to form F’· is knocked out by a single contrary instance. But I shall sometimes cite two or three such instances—for clarity’s sake and to provide practice in using the tables.

An example of exclusion or rejection of natures from the form of heat:

(1) reject: elemental nature because of the rays of the sun

(2) reject: heavenly nature because of ordinary fire, and especially underground fires, which are the most completely cut off from the rays of heavenly bodies

(3) reject: how fine-grained a body’s structure is because of the fact that all kinds of bodies (minerals, vegetables, skin of animals, water, oil, air, and so on) become warm simply by being close to a fire or other hot body

(4) reject: being attached to or mixed with another body that is hot because of the fact that red-hot iron and other metals give heat to other bodies without losing any of their own weight or substance

(5) reject: light and brightness because of boiling water and ·hot· air, and also metals and other solids that become hot but not enough to burn or glow

(6) reject: light and brightness because of the rays of the moon and other heavenly bodies (except the sun)

(7) reject: light and brightness because of the fact that red-hot iron has more heat and less brightness than the flame of alcohol

(8) reject: rarity because of very hot gold and other metals that have the greatest density

(9) reject: rarity because of air, which remains rare however cold it becomes

(10) reject: change in a body’s size or shape because of red-hot iron, which doesn’t become larger or change its shape

(11) reject: change in a body’s size or shape because of the fact that in thermometers, and the like, air expands without becoming noticeably warmer

(12) reject: destructive nature, or the forceful addition of any new nature because of the ease with which all bodies are heated without any destruction or noticeable alteration

(13) reject: expanding or contracting motion of the body as a whole because of the agreement and conformity of similar effects displayed by both heat and cold

(14) reject: the basic natures of things (as distinct from properties they have through antecedent causes) because of the creation of heat by rubbing things together There are other natures beside these; I’m not offering complete tables, but merely examples.

Not a single one of the ‘reject:’ natures belongs to the form of heat. In all our dealings with heat we can set those aside.

19. The process of exclusion is the foundation of true induction; but the induction isn’t completed until it arrives at something affirmative. Of course the excluding part ·of our work· is itself nothing like complete, and it can’t be so at the beginning. For exclusion is, obviously, the rejection of simple natures; so how can we do it accurately when we still don’t have sound and true notions of simple natures? Some of the notions that I have mentioned (such as the notions of elemental nature, heavenly nature and rarity) are vague and ill defined. I’m well aware of, and keep in mind, how great a work I am engaged in (namely making the human intellect a match for things and for nature); so I am not satisfied with what I have said up to here. I now go further, and devise and supply more powerful aids for the intellect—aids that I shall now present. In the interpretation of nature the mind should be thoroughly prepared and shaped up, so that it will at each stage settle for the degree of certainty that is appropriate there, while remembering (especially at the beginning) that the answer to ‘What is this that we have before us?’ depends to a great extent on what will come of it later on.

20. Truth emerges more quickly from error than from confusion, ·which implies that it can be worthwhile to aim for clarity even at the risk of going wrong·. So I think it will be useful, after making and weighing up three tables of first presentation (such as I have exhibited), to give the intellect permission to try for an interpretation of nature of the affirmative kind on the strength of the instances given in the tables and also of any others that may turn up elsewhere. I call this kind of attempt •‘permission for the intellect’ or •‘sketch of an interpretation’ or—·the label I shall actually use in this work·—•the ‘first harvest’.

A first harvest of the form of heat

Something that is perfectly clear from what I have said earlier should be borne in mind here, namely that the •form of a thing is present in each and every instance of the thing; otherwise it wouldn’t be its •form; from which it follows that there can’t be any counter-instances ·where the thing is present and the form isn’t·. Still, the form is much more conspicuous and obvious in •some instances than in others, namely in •those where the nature of the form is less restrained and obstructed and limited by other natures. Instances of •this kind I call ‘luminous’ or (·most of the time·) ‘revealing’ instances. So now let us proceed to the first harvest concerning the form of heat.

In each and every case of heat the cause of the nature of which heat is a special case appears to be motion. This shows most conspicuously in flames, which are on the move all the time, and in boiling or simmering liquids, which are also constantly in motion. It is also shown when motion stirs heat up or increases it—as happens with bellows and with wind (Third Table 29) and with other kinds of motion (28 and 31). It is also shown when fire and heat are extinguished by any strong compression, which checks and stops the motion (see 30 and 32). It is shown also by the fact that all bodies are destroyed or at any rate significantly changed by any fire or strong heat, which makes it quite clear that heat causes a tumult and agitation and lively motion in the internal parts of a body, which gradually moves it towards dissolution.

In certain cases heat generates motion and in certain cases motion generates heat, but that isn’t what I am saying when I say that motion is like a genus in relation to heat ·as one of its species·. What I mean is that heat itself is nothing but motion of a certain specific kind; I’ll tell you soon what special features of a case of motion make it qualify as a case of heat. Before coming to that, though I shall present three cautions that may be needed to avoid unclarity about some of the terms I shall be using.

·First caution: My topic is heat, not heat-as-we-feel-it·. Heat as we feel it is a relative thing—relative to humans, not to the world; and it is rightly regarded as merely the effect of heat on the animal spirits. Moreover, in itself it is variable, since a single body induces a perception of cold as well as of heat, depending on the condition of the senses. This is clear from the item 41 in the Third Table [here].

·Second caution: My topic is heat, not the passing on of heat·. Don’t confuse the form of heat with the passing on of heat from body to body, for heat is not the same as heating. Heat is produced by the motion of rubbing something that at first has no heat; and that’s enough to show that the transmission of heat is no part of the form of heat. And even when something is heated by another hot thing’s coming close to it, that doesn’t come from the form of heat; rather, it depends entirely on a higher and more general nature, namely the nature of assimilation or self-multiplication, a subject that needs to be investigated separately. [See here.]

·Third caution: My topic is heat, not fire·. Our notion of fire is a layman’s one, and is useless ·for scientific purposes·. What it counts as ‘fire’ is the combination of heat and brightness in a body, as in ordinary flame and bodies that are red hot. [Red-heat is treated as a kind of ‘burning’ in item 24 here.]

Having guarded against verbal misunderstandings, I now at last come to the true specific differences which qualify a case of •motion (·genus·) to count as a case of •heat (·species·).

The first difference then is this. Heat is an expansive motion in which a body tries expand to a greater size than it had before. We see this most clearly in flame, where the smoke or thick vapour obviously expands into flame.

It also appears in any boiling liquid, which can be seen to swell, rise and bubble, and goes on expanding itself until it turns into a body that is far bigger than the liquid itself, namely into steam, smoke, or air.

It appears also in all wood and ·other· flammable things, where there is sometimes sweating and always evaporation.

It is shown also in the melting of metals. Because they are highly compact, metals don’t easily expand and dilate; but their spirit expands, and wants to expand further; so it forces and agitates the lumpier parts into a liquid state. If the metal becomes hotter still, it dissolves and turns much of itself into a volatile substance.

It appears also in iron or rocks: they don’t liquefy or run together, but they become soft. Similarly with wooden sticks, which become flexible when slightly heated in hot ashes.

But this kind of motion is best seen in air, which a little heat causes to expand—see Third Table 38 [here].

It shows up also in the contrary nature, namely cold. For cold contracts all bodies—makes them shrink—so that in a hard frost nails fall out of walls, bronze vessels crack, and heated glass when exposed to cold cracks and breaks. Similarly, a little cooling makes air contract, as in 38. But I’ll say more about this when I deal properly with cold.

It’s no wonder that heat and cold should exhibit many actions in common (for which see the Second Table 32). This first specific difference ·helping to denarcate the species heat within the genus motion· concerns a feature of heat that is diametrically opposite to a feature of cold, because whereas heat expands cold contracts; but the third and fourth differences (still to come) belong to the natures both of heat and of cold.

The second difference is a special case of the first, namely: Heat is a motion in which the hot body •expands while it •rises. This is a case of mixed motion, of which there are many—e.g. an arrow or javelin •rotates while it •flies forward. Similarly the motion of heat is an expansion as well as a movement upwards.

This difference appears when you put a poker into a fire. If you put it in upright and hold it by the top, it soon burns your hand; if you put it in at the side or from below, it takes longer to burn your hand.

It can also be seen in fractional distillation, which men use for ·extracting essences from· delicate flowers that soon lose their scent. It has been found in practice that one should place the fire not below ·the distilling retort· but above it, so as to burn less. For all heat, not only flame, tends upward.

This should be tried out on the opposite nature, cold, to learn whether cold contracts a body downward as heat expands it upward. Here’s how to do it. Take two iron rods or glass tubes of exactly the same dimensions, warm them a little and place a sponge steeped in cold water or snow at the bottom of the one, and a similar one at the top of the other. I think that the end of the rod that has snow at the top will cool sooner than the end of the rod with snow at the bottom—the opposite of what happens with heat.

The third specific difference is this: heat is a motion that isn’t expansive uniformly through the whole ·hot· body, but only through its smaller particles; and this expansion ·in any one particle· is at the same time checked, repelled, and beaten back ·by the expansions of other particles·, so that there’s a back-and-forth motion within the body, which is irritated by all the quivering, straining and struggling that goes on; and from that comes the fury of fire and heat.

This ·specific· difference is most apparent in flames and in boiling liquids, where there are continual little rises and falls across their surface.

It also shows up in bodies that are so compact that when heated or ignited they don’t swell or expand in bulk—e.g. in red-hot iron, in which the heat is very sharp.

And it is apparent in hearth fires, which burn brightest in the coldest weather.

It also shows in the fact that when the air in a calendar glass [see item 38 here] expands without obstacles or counter-pressures, and thus expands at the same rate throughout, there is no perceptible heat. Also when an enclosed body of ·compressed· air escapes, no great heat is observed; that is because although the air bursts out with the greatest force, its only expansive motion is a motion of the whole, with no back-and-forth motions in the particles. . . .

It is also shown in this, that all burning acts on minute pores in the body in question, so that burning digs into the body, penetrating and pricking and stinging it like the points of countless needles. . . .

And this third specific difference is shared with the nature of cold. For in cold the contractive motion is checked by a tendency to expand, just as in heat the expansive motion is checked by a tendency to contract. Thus, whether the particles of a body work inward or outward, the mode of action is the same though the degree of strength may be very different; because on the surface of the earth we don’t have anything that is intensely cold. [See item (3) here.]

The fourth specific difference is a special case of the third. It is that the motion of pricking and penetrating must be fairly fast, not sluggish, and must go by particles—very small ones but a bit bigger than the smallest.

This difference is apparent when you compare the effects of •fire with the effects of •time or age. Age or time makes things wither, consumes and undermines them, reduces them to ashes, just as much as fire does, though it acts on even smaller particles than fire acts on; because that motion is very slow and acts on very tiny particles, there is no detectable heat.

It is also shown by comparing the dissolution ·in acids· of iron and gold. Gold is dissolved without any heat being stirred up, whereas iron, when it is dissolved about as quickly as gold, starts up a violent heat. This is because the solvent for gold enters the gold gently and works at a level of very small particles, so that the particles of the gold give way easily; whereas the solvent for iron enters the iron roughly and forcibly, and the particles of the iron are more stubborn.

It is also apparent in some gangrenes and cases of rotting flesh, which don’t arouse much heat or pain because the rotting process operates at the level of such tiny particles.

I offer this as the •first harvest—or •sketch of an interpretation—concerning the form of heat, made by way of •permission to the intellect [these three labels are introduced in 20 here.].

The form or true definition of heat can be derived from this first harvest. (I’m talking about heat considered absolutely, not heat relative to the senses.) Here it is, briefly:

•Heat is an expansive motion that is resisted, and that fights its way through the smaller particles ·of the hot body·.

Special case of this expansion:

•While expanding in all directions ·the hot body· has a tendency to rise.

Special case of the struggle through the particles:

•It is not very slow; rather it is fast and has some force.

This tells us how in practice to create heat. Here is the story:

In some natural body, arouse a motion to expand; and repress this motion and turn it back on itself so that the expansion doesn’t proceed evenly, but partly succeeds and is partly held back.

If you do that you will undoubtedly generate heat. It makes no difference whether

•the body is made of earthly elements or contains heavenly substances,
•is luminous or opaque,
•is rare or dense,
•is spatially expanded or still of its original size,
•tends towards dissolution or keeps its original condition,
•is animal, vegetable, or mineral (water, oil or air),

or any other substance that is capable of the motion described. Sensible heat is the same, but considered with reference to the senses. Let us now proceed to further aids.

[That last remark refers to the ‘aids’ that were promised in 19 here; the first such ‘aid’ has been 20. A reminder about ‘the tables of first presentation’:

•the first table, of essence and presence, starts here;
•the second table, of divergence or nearby absence, starts here;
•the third table, of degrees or of comparison, starts here;
•‘the table of exclusion or rejection’ starts here;
•‘the first harvest’ starts here.

This reminder may be useful as a guide to Bacon’s next remark.]

21. So much for the tables of •first presentation and of •rejection or exclusion, and the •first harvest based on them. Now we have to proceed to the other aids to the intellect in the interpretation of nature and in true and perfect induction. I’ll present them in terms of heat and cold whenever tables are appropriate; but when only a few examples are needed I’ll take them from all over the place, so as to give my doctrine as much scope as possible without creating confusion.

[We are about to meet the phrase ‘privileged instances’. The Latin praerogativa instantarum strictly means ‘privilege of instances’, but Bacon always handles it as though it stood for a kind of instance, not a kind of privilege. The use of ‘privilege’ to translate praerogativa is due to Silverthorne, who relates it to the centuria praerogativa in ancient republican Rome—the aristocrats’ privilege of voting first and thus having the best chance to influence the votes of others.]

My topics will be, in this order:

1. privileged instances

2. supports for induction

3. the correcting of induction

4. adapting the investigation to the nature of the subject

5. which natures should be investigated first, and which later

6. the limits of investigation, or a synopsis of all natures in the universe

7. practical consequences

8. preparations for investigation

9. the ascending and descending scale of axioms.

[There are twenty-seven classes of privileged instances, some with a number of sub-classes. Bacon’s discussion of them runs to the end of the work. The other eight topics were to have been dealt with in later instalments of the Great Fresh Start, which he never wrote.]

22. Class 1 of privileged instances: solitary instances. Those are ones in which the nature we are investigating

appears in things that have nothing else in common with other things that have that nature,

or ones in which the nature we are investigating

does not appear in things that have everything else in common with other things that do have that nature.

·I put these first · because it is clear that they save us from detours, leading quickly and securely to exclusions, so that a few solitary instances are as good as many.

Suppose for example that we are investigating the nature of colour: in that context prisms, crystals, dew-drops and the like, which make colours in themselves and project them outside themselves onto a wall, are solitary instances. For they have nothing else in common with the colours inherent in flowers, coloured stones, metals, woods, etc.—i.e. nothing but colour. From which we can easily draw the conclusion that colour is merely a modification of the light that the object takes in. With prisms, crystals etc. the light is modified by the different angles at which the light strikes the body; with flowers, coloured stones etc. it is modified by various textures and microstructures of the body. These instances are •resemblance-solitary.

In that same investigation of light: the distinct veins of white and black in a piece or marble, and the variegation of colour in flowers of the same species, are solitary instances. The black and white streaks in marble have almost everything in common except their colour, and so do the streaks of pink and white in a carnation. From this we can easily infer that colour doesn’t have much to do with the intrinsic nature—·the microscopic fine texture·—of a body, but only on the quasi-mechanical arrangement of its larger parts. These instances are •difference-solitary. . . .

23. Class 2 of privileged instances: shifting instances. Those are ones where the nature under study is •shifting towards being produced when it didn’t previously exist, or •shifting towards non-existence when it existed before. Shifting instances, whichever kind of shift they involve, are always twofold, or rather it is one instance in which the movement is continued until it reaches the opposite state.

[At this point some material is removed, and will be reinserted as a paragraph between *asterisks* below; it is easier to understand there than it would be here.]

Here is an example of a shifting instance. Suppose we are investigating whiteness: shifting instances in which the shift is towards production or existence ·of whiteness· are

unbroken glass shifting to powdered glass ordinary water shifting to water shaken up to make foam.

Plain glass and water are transparent, not white, whereas pounded glass and foaming water are white, not transparent. So we have to ask what happened to the glass or water in this shift. Obviously, the form of whiteness is brought in by the pounding of the glass and the shaking of the water; but we find that nothing has occurred except the breaking up of the glass and water into small parts, and the introduction of air. So we have this result:

Two bodies, air and water (or: air and glass) which are more or less transparent come to exhibit whiteness as soon as they are broken up into small bits ·and the bits are mixed·, this whiteness being brought about by the unequal refraction of the rays of light.

This is a big step towards discovering the form of whiteness.

*Such instances don’t just lead quickly and securely to exclusions, but also narrow down the search for the affirmation or the form itself [‘exclusion’ and ‘affirmation’ are introduced in 15 here]. For the form of a thing must be something that is introduced by a shift, or removed and wiped out by a shift in the other direction. Of course every exclusion supports some affirmation, but the support is more direct when the exclusion comes from one case rather than from a number of cases. And my discussion has made it clear that the form that comes to light in a single instance leads the way to the discovery of it in all the rest. And the simpler the shift, the more value we should attach to the instance. And another thing: shifting instances are of great value in the practical part ·of scientific inquiry·: a shifting instance exhibits •the form ·under investigation· linked with •the cause of its existing (or the cause of its not existing); that provides great clarity in one instance and an easy transition to others. But shifting instances create a certain danger against which I should warn you: they may lead us to link the form too closely to its efficient cause, and so encourage a false view of the form, drawn from a view of the efficient cause. The efficient cause is always understood to be merely the vehicle for or bearer of the form. It is not hard to avoid this danger in a properly conducted exclusion.*

I should give an example of this danger. A mind that is led astray by efficient causes of this sort will too easily conclude that •air is always required for the form of whiteness, or that •whiteness is generated only by transparent bodies—both of which are entirely false, and refuted by numerous exclusions. What will be found (setting air and the like aside) is this:

all the particles that affect vision are equal
unequal and simply textured
unequal with complex regular texture
any but black
unequal and complex in an irregular way

So now we have before us an instance with a shift to the •production of the nature under study, namely whiteness. For an instance that shifts to the •destruction of the same nature of whiteness, consider breaking up foam or melting snow. In each case, what you then have is water, not broken into little particles and not mixed with air, and this sheds whiteness and puts on transparency.

It’s important to note that shifting instances include not only those in which the nature under study shifts toward production or toward destruction, but also those in which the nature shifts towards increasing or decreasing. It’s because these also contribute to revealing the form, as can be clearly seen from the definition of form that I have given ·in 17·, and the Table of Degrees [starting here]. Paper that is white when dry become less white and nearer to being transparent when it is wetted—i.e. when air is excluded and water introduced. The explanation of what is happening here is analogous to the explanation of the first shifting instances.

24. Class 3 of privileged instances: revealing instances, which I have already mentioned in the first harvest concerning heat, and which I also call ‘luminous’ and ‘freed and predominant’. They are the instances in which the nature under study is revealed

naked and standing on its own feet, and also
at its height and in full strength,

not muffled by any impediments. This is either because •there aren’t any impediments in this instance or because •there are some but the nature we are studying is present in such strength that it holds them down and pushes them around. ·Here is the background setting for these revealing instances·:

Every body is capable of having many forms or natures linked together; they can crush, depress, break and bind one another so that the individual forms are obscured. But we find that in some subjects the nature under investigation stands out from the others, either because there are no obstacles or because its vigorous strength makes it prominent.

Instances of this kind reveal the form with special clarity.

But we should be careful in our handling of ·what seem to be· revealing instances, not rushing to conclusions. When something reveals a form very conspicuously and seems to force it on the notice of our intellect, we should view it with suspicion and should avail ourselves of a strict and careful exclusion ·of other potentially relevant features, rather than abruptly brushing them aside in our enthusiasm for the conspicuous nature that has attracted our attention·.

Suppose, for example, that we are investigating the nature of heat. As I said earlier [in item 38 here], the motion of expansion is the main element in the form of heat, and a revealing instance of that is a •thermometer. Although •flame obviously exhibits expansion, it doesn’t show expansion as an ongoing process, because a flame can be so quickly snuffed out. Nor does •boiling water provide a good display of expansion in its own body ·as water· because it so easily turns into vapour or air. As for red-hot iron and its like: they are so far from exhibiting expansion as an ongoing process that their expansion is almost imperceptible; that’s because their spirit is being crushed and broken by the coarse and compact particles, which curb and subdue the expansion. But a thermometer clearly displays expansion in air, revealing it as conspicuous, progressive, and enduring rather than transitory.

To take another example: suppose the nature inquired into is weight. A revealing instance of weight is mercury. It is heavier than anything else except gold, which is only slightly heavier; and mercury does a better job of indicating the form of weight than gold does, because gold is solid and compact—features that seem to come from its density— whereas mercury is liquid and full of spirit despite being much heavier than the diamond and other bodies that are thought to be the most solid. This reveals that the form of heaviness or weight depends simply on the quantity of matter and not on how compact the body is.

25. Class 4 of privileged instances: concealed instances, which I also ·though not again in this work· call ‘instances of the twilight’. They are pretty nearly the exact opposites of revealing instances. They exhibit the nature under investigation at its lowest strength, as though it were in its cradle, newly born, making its first attempts but buried under and subdued by a contrary nature. Still, such instances are very helpful in the discovery of forms; because just as

revealing instances lead easily to •specific differences,

so also

concealed instances are the best guides to •genera,

i.e. to the common natures of which the natures under investigation are merely special cases. ·That is to say, revealing instances help us to move down the classificatory table, concealed instances help us to move up·.

Suppose for example that the nature under investigation is •solidity or a thing’s holding its shape, the opposite of which is •fluidity. Concealed instances of this are ones that exhibit some low level of shape-holding in a fluid—for example a bubble of water, which has a sort of shaped skin made of water. Similarly with trickling water: if the water keeps coming, the drops lengthen themselves out into a thin thread so as to keep the stream unbroken; and if there isn’t enough water for that, the water falls in round drops, that being the shape that best preserves the water from breaking up ·into still smaller portions·. But the instant the thread of water stops and the drops begin, the water jumps back upwards so as to avoid breaking. And in metals, which when melted form thick fluids, the molten drops often jump back up and stay there. . . . The same kind of thing can be seen in the children’s game when they take water, thicken it a little with soap, and blow it through a hollow reed: this combines the water with air so as to make a cluster of bubbles that is firm enough to be thrown some distance without breaking up. But foam and snow provide the best examples of this phenomenon. They become almost solid enough to be cut with a knife, although they are made out of two fluids—air and water. All of this pretty clearly indicates •that ‘solid’ and ‘liquid’ are ·not useful terms in the present context, because they are· layman’s notions which relate ·not to the scientific facts about a thing but only to how it strikes· our senses. It also indicates •that in fact all bodies have a tendency to avoid being broken up, a tendency that is weak in homogeneous bodies (which is what fluids are), and stronger in bodies made up of different kinds of materials (·the ones the layman calls ‘solid’·). That is because a body is bound together when heterogeneous matter is introduced to it, whereas the insertion of homogeneous matter dissolves the body and makes it fall apart.

Here are three more examples. (1) Suppose that the nature we are investigating is the attraction or coming together of bodies. The best revealing instance of the form of this is the magnet. There is also the non-attracting nature—the contrary of the attracting one—and this can even be found in the same substance. Thus iron doesn’t attract iron, lead doesn’t attract lead, or wood wood, or water water.

[In what follows, an ‘armed’ magnet is one equipped with an ‘armature’ in the sense of ‘a piece of soft iron placed in contact with the poles of the magnet, which preserves and increases the magnetic power; or any arrangement which produces the same result’ (OED). Another such arrangement is an ‘armature’ in our sense of the word—coils of wire conducting electricity— but that wasn’t discovered as a means of magnetism until two centuries later.]

Now a concealed instance ·of attraction· is provided by •a magnet armed with iron, or rather by •the iron in an armed magnet. Its nature is such that

an armed magnet does not attract iron from a distance more powerfully than an unarmed magnet does,


when the iron in an armed magnet touches some other iron, the magnet supports a far greater weight of iron than a simple unarmed magnet would.

This is because of the similarity of substances, iron on iron—an effect that was latent in the iron ·all along·, but was completely concealed before the magnet was brought into play. So it is clear that the form of coming-together is something that is lively and strong in the magnet, feeble and latent in iron. (2) It has been noticed that small wooden arrows with no iron points, shot from large guns into the sides of ships or into other wooden targets, penetrate more deeply than they would if they were tipped with iron. This is because of the similarity of substances, wood on wood, although this property had previously been latent in the wood—·only latent, and thus concealed·. (3) Similarly, whole bodies of air (water) don’t obviously attract other bodies of air (water), but the likelihood of a bubble’s bursting is increased when it is touched by another bubble. This is because of water’s ·usually concealed· inclination to join with water, and air’s to join with air. Such concealed instances (which are very useful, as I have said) show up most conspicuously in small portions of bodies. The reason for that is that larger masses follow more general forms, as I’ll explain in due course.