Studies in History and Philosophy of Science 51 (2015) 1e10

Contents lists available at ScienceDirect

Studies in History and Philosophy of Science

journal homepage: www.elsevier.com/locate/shpsa

Qualitative novelty in seventeenth-century science: Hydrostatics from

Stevin to Pascal
Alan F. Chalmers
Unit for History and Philosophy of Science, University of Sydney, Carslaw Building F07, NSW 2006, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Two works on hydrostatics, by Simon Stevin in 1586 and by Blaise Pascal in 1654, are analysed and
Received 25 August 2014 compared. The contrast between the two serves to highlight aspects of the qualitative novelty involved in
Received in revised form changes within science in the first half of the seventeenth century. Stevin attempted to derive his theory
18 December 2014
from unproblematic postulates drawn from common sense but failed to achieve his goal insofar as he
Available online
needed to incorporate assumptions involved in his engineering practice but not sanctioned by his
postulates. Pascal’s theory went beyond common sense by introducing a novel concept, pressure.
Keywords:
Theoretical reflection on novel experiments was involved in the construction of the new concept and
Hydrostatics;
Pressure;
experiment also provided important evidence for the theory that deployed it. The new experimental
Simon Stevin; reasoning was qualitatively different from the Euclidean style of reasoning adopted by Stevin. The fact
Blaise Pascal; that a conceptualization of a technical sense of pressure adequate for hydrostatics was far from obvious is
Experiment; evident from the work of those, such as Galileo and Descartes, who did not make significant moves in
Mechanism that direction.
Ó 2015 Elsevier Ltd. All rights reserved.

When citing this paper, please use the full journal title Studies in History and Philosophy of Science

1. Introduction the consequences of Pascal’s hydrostatics can be seen as conse-

In this paper I aim to shed light on the nature of the changes that
took place in science in the first half of the seventeenth century by
looking at the move beyond the science of weight to include hy-
drostatics. In particular, I compare two texts on hydrostatics
composed sixty eight years apart, The Elements of Hydrostatics,
written by Simon Stevin in 1586 and The Equilibrium of Liquids, the
first of two treatises written by Blaise Pascal in 1654 and published
posthumously in 1663, the second one being concerned with
pneumatics rather than hydrostatics.1
Some immediately apparent features of the two texts signal the
need for a discerning reading of the situation. Stevin’s text is
somewhat alien to a reader trained in physics, appearing as one
more akin to Euclidean geometry than empirical science. By
contrast, Pascal’s text would not be out of place in a modern course
on undergraduate physics. However, it is also the case that most of
E-mail address: achalmers@usyd.edu.au.
quences of Stevin’s version or some modest extension of it. It may
be tempting to take Pascal’s rhetoric at face value and see the
emphasis on experimental support as the novel feature of his
approach. There are two prima facie difficulties to be confronted
here. Firstly, many of the experiments described by Pascal are
modifications of situations described by Stevin. Secondly, it is very
likely that Pascal did not in fact perform many of the experiments
described in his text. Such conundrums help set the scene for my
exploration of the case.
I end this introduction by foreshadowing the conclusions I will
reach. Stevin interpreted geometry and the science of weight as a
body of theorems deduced from unproblematic postulates. To
extend this approach to hydrostatics Stevin needed postulates that,
on the one hand, had sufficient content to yield theorems consti-
tuting the new science and, on the other hand, were sufficiently
unproblematic to be granted at the outset. Because significant hy-
drostatic phenomena transcended and posed problems for the
common sense knowledge of the time, Stevin was unable to satisfy
this demand. Stevin’s presentation of his hydrostatics in Euclidean

http://dx.doi.org/10.1016/j.shpsa.2015.01.001
0039-3681/Ó 2015 Elsevier Ltd. All rights reserved.
2 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10
guise masked the extent to which he introduced hydrostatic
knowledge into his theory that went beyond what was licensed by
his postulates. Stevin did not succeed in achieving a theoretical
grasp of the difference between liquids and solids that would make
possible a science of hydrostatics that went beyond the science of
weight. In their dealings with hydrostatics, Galileo, Beeckman and
Descartes met with no more success than Stevin in solving this
problem. It was not until Pascal that the move was made to a
distinction between solids and liquids via the addition of the
concept of pressure to the concept of weight. Because the new
concept moved beyond common sense, it became necessary to
recognise that the case for the new hydrostatics was essentially
empirical or experimental rather than relying on proofs from pre-
given, unproblematic postulates, Euclidean style.
2. The background to Stevin’s hydrostatics
There is a notion of mechanism that is as old as civilisation itself.
Levers, slings and tow ropes are mechanisms in the sense I have in
mind. By Stevin’s time clocks were familiar enough to serve as
archetypal mechanisms. Characteristic of such mechanisms is the
way in which a cause brings about its effect by way of the pushes
and pulls acting between neighbouring parts of the system that
links them. In the following I refer to such mechanisms as clock-
work mechanisms and to explanations that invoke them as clock-
work explanations. Clockwork explanations are intelligible in a
common sense and are useful. They facilitate purposeful manipu-
lation of the material world.
Clockwork mechanisms were theorised in what I will refer to as
the science of weight. That science, having origins in Pseudo-
Aristotle and Archimedes, encompassed balances, levers, pulleys
and the like and had been developed with sophistication and in
detail by the end of the sixteenth century. Not only did this ‘science’
provide truths that could be taken for granted, such as the law of
the lever, but it also served as a model of a theorised, mathematized
body of knowledge.
Archimedes had moved a little beyond the science of weight to
formulate the beginnings of hydrostatics in his work on floating
bodies. A solid immersed in a liquid experiences an upthrust equal
to the weight of a mass of the liquid equal in volume to that of the
immersed solid. As we shall see, Stevin was able, not only to take
advantage of the substance of Archimedes work on floating bodies,
but also to adopt its Euclidean style.
A scholar aiming to develop a science of hydrostatics in the late
sixteenth century did not need to invent a distinction between
solids and liquids any more than Archimedes discovered the phe-
nomenon of floating. Like a working knowledge of weight, a grasp
of the distinction between liquids and solids was incorporated into
everyday life and into technologies long before precise and math-
ematized versions of such notions were fashioned. Liquids flow
whereas solids do not. As a result, solids sink or float in liquids, but
not in other solids. If water is transmitted from a lake at high alti-
tude via a pipe then it is as liable to leak from the top as the bottom
of the pipe, and the flow will not be halted if the pipe needs to rise
over a subsidiary hill whose maximum height lies beneath the lake.
Neither of these qualitative facts is true in the case of pipes trans-
ferring sand or gravel from a high to a low altitude, and sand or
gravel ejected from such pipes will form a pile in a way that liquids
will not. If a perforated bladder filled with water is squeezed, the
water is ejected in all directions, and not only in the direction of the
squeeze. Water finds its own level.
Some common hydrostatic phenomena could be seen as prob-
lematic, as indeed they are if one’s thinking is confined to weight.
Two columns of liquid with different diameters and in communi-
cation via their base will rest in equilibrium with their heights level,
raising the puzzle of how the lighter amount of liquid in the nar-
rower tube can support the heavier amount in the other.2
I complete these introductory remarks on the background to the
formalization of hydrostatics with some reflections on use of the
term ‘pressure’. The term had a wide range of uses long before the
modern technical sense of it was fashioned in the seventeenth
century. The Latin terms pressio/pressionem stemming from the
verb premo, to press, were used in a variety of senses that overlap
with everyday usage of ‘pressure’ in the seventeenth century and in
modern times. The most common usage involved the forces
resulting from weights bearing down on surfaces but extended
more widely to include the results of various kinds of pressing, such
as that used to mould clay or to force contents into a container.
Insofar as the common sense of the term pressure was linked to
pressing, it suggested a directed force, whereas pressure in the
technical sense is undirected. It is a scalar not a vector. Our analysis
must clearly probe deeper than the mere identification of usage of
the term. As a matter of fact, the noun ‘pressure’ is not used either
by Stevin (druck) or by Pascal (la pression) in their respective
treatises on hydrostatics although they do talk freely about press-
ing. The English translations of the two treatises both freely employ
the term ‘pressure’ and so raise the danger of reading more into the
texts than is warranted.
3. Simon Stevin and The Art of Weighing
Stevin was one of that new breed of mathematically trained
engineers who made contributions to knowledge construction
outside of a university context. He was actively involved in the
renewed interest in mathematics fed by the increased availability of
Ancient sources, especially the works of Archimedes, and was
actively involved in responses to the practical demands of a rapidly
changing society. Soon after matriculating at the University of
Leyden in 1583 he was publishing original works in arithmetic and
geometry whilst making his way as an engineer. During the 1580s
he had been granted patents for various devices, a number of them
connected with drainage and dredging, and formed a business
partnership with his friend Johan Cornets de Groot in order to put
his inventions to practical use. In the early 1590s he entered the
service of Prince Maurice of Nasssau, acting as his tutor and also
advisor on various practical matters. Most of Stevin’s publications
after his association with Prince Maurice were written in the form
of textbooks for the edification of his patron and student. They
covered a range of areas including astronomy, navigation, military
science, architecture, music and optics. These later works involved
an opportunistic and far from uniform combination of empirical
and mathematical considerations which contrasts with the strict
mathematical character of his earlier works. The two books by
Stevin which are relevant for the concerns of this paper, The Art of
Weighing and The Elements of Hydrostatics, were presented as
mathematical works in the style of Euclid and Archimedes. They
were published in 1586, more than half a decade before Stevin’s
professional association with Prince Maurice began.
The Art of Weighing is presented as a body of theorems derived
from postulates. The latter include idealizations, such as the as-
sumptions that the arms of a balance are inflexible and that verti-
cals are parallel, and also assumptions with empirical content, such
as the postulate that equal weights hung from equal balance arms
will be in equilibrium. The subsequent theory, outlined in Book 1 of
the work, extends treatment of vertically acting weights to non-
vertical actions mediated by inclined planes and pulleys. Book 2
is concerned with the calculation of centres of gravity of a range of
plane and solid figures. The mathematical techniques employed
include splitting the action of a weight into its components, adding
weights using the parallelogram of forces, and locating centres of
 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10
gravity by appealing to limits which usage marks a step towards the
infinitesimal calculus.
Stevin followed his treatise on The Art of Weighing with a sequel
on The Practice of Weighing. This included practical techniques for
the determination of centres of gravity and of measuring weights.
He also described various machines employing ingenious ar-
rangements of levers and pulleys, including an exceedingly
powerful machine that he called ‘The Almighty’, involving a wheel
and axle driven by a crank handle via a system of gear wheels.
Insofar as this machine was capable of hauling heavy ships over
sand, it extended the notion of a machine beyond the ability to raise
weights to that of overcoming other resisting forces.
Stevin’s Art of Weighing was intended to conform to what I will
call the Euclidean ideal. According to that ideal, a body of knowledge
is theorised or mathematized to the extent that it can be repre-
sented as a body of theorems that are the deductive consequences
of postulates that are sufficiently unproblematic to be granted at
the outset. Stevin explicitly likened his theory to Euclidean geom-
etry (Dijksterhuis, 1955, p. 97). The truth of a theorem in geometry,
such as Pythagoras’s theorem, may not be at all obvious but can be
shown to be true nevertheless by deriving it from postulates that
are obvious or unproblematic. The same can be said of theorems in
the Art of Weighing as Stevin made explicit.
Since some matters of an elementary nature are common
knowledge, and need not be proved, while other matters of a
more veiled character might give the critic cause to criticize that
which does not deserve criticism, we shall, after the manner of
mathematicians, before arriving at the propositions, postulate
that the following things be granted (Dijksterhuis, 1955, p. 111).
The grounds for the theorems constituting the Art of Weighing lie in
the fact that they are deductive consequences of unproblematic
postulates. Given such grounding, the theorems can be applied to a
range of practical situations including novel ones. The situations
described in the Practice of Weighing are presented as applications
rather than tests of the theory.
Given this background we can characterise the task that Stevin
set himself in compiling the Elements of Hydrostatics in the
following way: Stevin aspired to extend his Euclidean science of
weight to hydrostatics. To do so he needed to capture theoretically
what lies behind the common sense distinction between liquids
and solids. In the next section I offer a critical evaluation of Stevin’s
efforts in that regard.
4. Stevin’s hydrostatics
4.1. Introductory remarks
It is not over-generous to summarise the content of Stevin’s
hydrostatic theorems as follows:
Normal to any element of solid surface bounded by a liquid,
whatever its orientation, there is a force equal to the weight of a
column of liquid with cross-section equal to that of the surface
element in question and a height equal to the depth of that
surface element beneath the liquid surface.
This assertion, sometimes referred to as ‘Stevin’s law’, is true from a
modern point of view and, moreover, gives hydrostatics much of
what it needs. However, this circumstance notwithstanding, I will
argue below that Stevin’s case for his hydrostatics suffered serious
shortcomings that seem to have gone unnoticed, judging by the
secondary literature. Pierre Duhem (1905, p. 603), one of the few
historians to have paid detailed attention to Stevin’s hydrostatics,
may well have been justified in acclaiming the latter for its novelty,
but, as we shall see, his extension of that acclaim to include rigor
3
cannot be allowed. Dijksterhuis (1955, p. 377), another historian to
have given serious attention to Stevin’s hydrostatics, also failed to
pick up on the shortcomings that I will identify. The attribution of
rigour to the deductions in Stevin’s hydrostatics seems to have
remained unchallenged in more recent literature. For example, in
their otherwise excellent historical analysis of early work of Des-
cartes on hydrostatics Gaukroger and Schuster (2002, pp. 539e540)
impute to him a firm belief in Stevin’s rigour and seem to endorse
this judgement themselves, even though doing so is not material to
their historical case.
Given the nature of Stevin’s project it is pertinent to raise the
question of how the hydrostatic content of his theory enters in via
his postulates. The only significant candidate in this respect is
Stevin’s Postulate III. It reads as follows:
The weight causing a vessel to sink less deep to be lighter, but
the weight causing it to sink deeper to be heavier, and that
causing it to sink to the same depth, equally heavy (Dijksterhuis,
1955, p. 397).
I do not suggest there are grounds for having qualms about granting
Stevin this postulate. What I do wish to point out is that it does not
contain enough to form the basis of his hydrostatics. In particular,
insofar as it refers only to the downward action of weight and
vertical resistance to it, Postulate III cannot yield the isotropic hy-
drostatic forces involved in Stevin’s hydrostatics and implied by
‘Stevin’s law’. The thrust of my critique of Stevin’s theory is that in
deriving his theorems he introduces or implies key assumptions
that are neither included in nor are consequences of his postulates.
In the next two subsections I substantiate my case by scrutinizing
Stevin’s derivation of two key theorems that specify the force on a
horizontal and on a vertical plane, respectively.
4.2. The force on a horizontal plane
Stevin could cope with a number of otherwise problematic hy-
drostatic phenomena through his realisation that the force exerted
by a liquid on a horizontal surface, whether it presses on it from
above or below, depends only on the area of the surface and its
depth below the liquid surface and is otherwise independent of the
total amount of liquid. Stevin’s Proposition X, which expresses this
fact, is unproblematic in the case of water contained in a regular
cylindrical vessel where it implies that the force on the base of the
cylinder will be equal to the weight of the water. In other cases the
predictions of the theorem are by no means self-evident and to
some degree counter-intuitive. In the situation depicted in Fig. 1,
the theorem predicts the correct result that the force exerted on the
base CD by the small amount of water contained in EFBA will be
equal to the weight of a regular cylinder of water with base CD and
height, h. Consequently, it can take a weight of several hundreds of
pounds to support an ounce of water.
Stevin’s proof of his theorem utilises a thought experiment that
involves replacing volumes of liquid by equal volumes of a solid
whose density is equal to that of the liquid. In the case depicted in
Fig. 1, imagine that we start with a regular cylinder of water on the
base CD with height, h. We now imagine that all the water other than
the small amount contained in EFBA is replaced by a solid equal in
density to the water. Provided the introduced solid is freely floating
in the water, it is reasonable to suppose that the move will have no
effect on the force on CD. Indeed, this will be a consequence of
Stevin’s Postulate III. However, what Stevin needs to complete his
argument and arrive at the situation depicted in Fig. 1 is to move
beyond the situation where the introduced solid floats to one where
it is secured so as the form a solid container for the water CDBA.
Stevin needed to, and did, assume that this move will have no effect
on the force on the base CD either. This step in the argument is not
4 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10

Given a regular bottom [surface] whose highest point is in the

water’s surface: the weight resting against it is equal to half of
the prism of water whose base is equal to that bottom, and
whose height is the vertical from the highest point of the bottom
to the plane parallel to the horizon through the lowest point of
the bottom. (Dijksterhuis, 1955, p. 421.)
In the case depicted in Stevin’s diagram (Fig. 2) the force acting
horizontally on the surface ACDE exerted by the rectangular prism
of water AGFBDC is, according to Proposition XI, the weight of the
wedge of water ACHDE which has surface ACDE as base and a height
DH equal to CD, the depth of the lower edge of plane ACDE beneath
the water surface. The proof of this is complicated by the fact that
the degree to which the water presses on the vertical surface varies
with height, from zero at the level AC to a maximum at the level ED.
Stevin was able to rise to the challenge posed by this complication.
Stevin calculates the force on the vertical plane by trapping it
between upper and lower limits that he can show converge on each
other, a technique that Galileo and Descartes were to use in the
context of free fall several decades later. Stevin begins by dividing
the prism of water into four slabs of equal height by equally spaced
horizontal planes through RV, SX and TY, so that the vertical plane
ACDE is divided into four equal rectangles. To calculate a lower limit
to the total force on ACDE Stevin assumes that the degree to which
the water presses horizontally is constant across each of the four
rectangles and is equal to the degree of pressing at the upper level
Fig. 1. A hydrostatic paradox. It can take a weight of many pounds to support an ounce of each rectangle. By adding the forces on each of the four rectan-
of water. gles calculated in this way Stevin gets a total that is less than the
force actually acting on ACDE. Stevin now repeats the process by
acknowledged by Stevin and, what is more, it is problematic insofar
assuming, once again, that the degree of pressing is constant across
as it goes beyond what is entailed by his postulates.
each rectangle, but this time it is assumed to be equal to the degree
The argument, that replacing a volume of water by the same
of pressing at the lowest level of each of them. The sum of the four
volume of a solid of equal density results in no change in the force on
forces gives a result that is greater than the force on ACDE. Stevin
the base of a container, is persuasive so long as the solid is floating
now imagines the number of divisions is increased from 4, to 8, 16
freely in the water. Once the solid is attached to the sides of the
and so on and demonstrates that the upper and lower limits to the
original container to form one of a different configuration the situ-
force on ACDE converge on each other, yielding a result for the force
ation is significantly different. A solid attached to the side of a
on ACDE that is in conformity with Proposition XI.
container, unlike a freely floating one, does not press down on the
We should not be misled by Stevin’s mathematical ingenuity nor
base of the container at all. What is at issue here is a fundamental
by the fact that the result he arrives at is correct from a modern
difference between the behaviour of solids and liquids that Stevin has
point of view. There are two indispensable assumptions implied in
in effect assumed without acknowledgement or justification. If a
Stevin’s argument which he does not identify as assumptions and
solid rests, and weighs down, on a table the table acts on the solid
which, moreover, are neither included in nor implied by his pos-
with a force of reaction equal to the weight of the solid, resulting in
tulates. Stevin assumes that the rectangular prism of water does
equilibrium. If the weight is now supported by a clamp so that it no
exert a horizontal force on a vertical plane and he further assumes
longer weighs down on the table the reaction of the table on the solid
that the force acting horizontally on a vertical surface at some level
becomes zero. Suppose, by contrast, our solid weighs down on a
is equal to the force acting vertically on a horizontal surface of the
liquid denser than it. Its weight will be countered by an upthrust from
the liquid resulting in equilibrium. Suppose that the floating solid is
now clamped so that it no longer weighs down. In this case the re-
action of the liquid does not fall to zero. The liquid will continue to
press on the solid as before. Stevin’s failure to distinguish between a
body floating in water and a supported body resting against water
results in an argument that lacks deductive rigor unless assumptions
not included in or implied by Stevin’s postulates are added.
The precise nature of Stevin’s argument here is not totally
transparent. Fortunately, I do not need to rest my overall case on my
interpretation of it. When it comes to Stevin’s derivation of his
Proposition XI, the force on a vertical plane, the inadequacy of his
deduction is blatant and unavoidable, notwithstanding the fact that
the inadequacy has been overlooked in the literature hitherto. I
analyse Stevin’s argument in the following Section.

4.3. The force on a vertical plane

Stevin’s Proposition XI specifies the force on a vertical plane. It Fig. 2. Stevin’s diagram relating to the force of water pressing against a vertical plane.
reads as follows: From Dijksterhuis (1955), facing p. 423.
 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10
same area at that level. This latter force is specified by Proposition X
which Stevin was able to invoke. Stevin’s postulates say nothing
about horizontally acting forces. Postulate III, which I identified
above as the only serious contender for the point at which signif-
icant content enters into Stevin’s hydrostatics, refers only to
weights acting vertically downwards and the vertically acting
resistance they encounter. Stevin’s expression for the force on a
vertical plane did not follow from his postulates any more than his
expression for the force on a horizontal plane did.
4.4. The significance of the invalidity of Stevin’s arguments
In this section I spell out the significance of the lapses in
deductive adequacy that I have identified in Stevin’s hydrostatics,
insisting that it goes far beyond mere logical quibbles.
My first point is that when Stevin introduced into his arguments
assumptions not included in or implied by his postulates he was
violating the standards set down in the Euclidean ideal that he
himself invoked. The demand that his theorems be derived from
postulates sufficiently unproblematic to be granted at the outset is
Stevin’s, not mine, and his introduction of assumptions not licensed
by his postulates violated that demand.
My second, more significant, point is that the introduction of
unwarranted assumptions by Stevin occurs in precisely those cir-
cumstances in which the hydrostatic phenomena in question
appear problematic from a point of view restricted to consider-
ations of weight. How can it be that it takes a weight of many
pounds to support an ounce of water? How can a body of water
press horizontally given that the horizontal component of weight is
zero? Given that it does press horizontally, how can it be that the
force on a lock gate depends only on the height of the water and not
on its extent? Surely these phenomena are ‘matters of a more
veiled character’ that ‘might give the critic cause to criticise what
does not deserve criticism’ and which Stevin must seek to clarify by
deriving the relevant phenomena from ‘common knowledge that
need not be proved’.3 There is no doubt that the validity of the
assumptions that Stevin did covertly insert into his hydrostatics
were familiar to him in the context of his hydrostatic practice.
Stevin, a hydrostatic engineer in the Netherlands, could hardly have
been unaware that water presses horizontally against a dyke or a
lock gate. The issue at stake concerns how such knowledge was
comprehended and theorised in Stevin’s hydrostatics. If there is a
puzzle posed by the fact that water presses horizontally against a
lock gate and to a degree that is independent of the extent of the
water, it does not help, given Stevin’s aspirations, to add such
knowledge to his postulates as assumptions to be granted at the
outset.
Stevin did raise the issue of the force on a lock gate in the
Practice of Hydrostatics.
As regards the 11th proposition, from it is evident, among other
things, what is the weight of the water pressing against either
side of the gate of a lock and the like. Also, that the water on one
side, even if it were only the width of a straw, exerts the same
force against it as the waters having the breadth of the ocean on
the other side, provided they are at the same level (Dijksterhuis,
1955, p. 497).
It is true that the force on a lock gate follows from Proposition XI,
but it does so only because the relevant knowledge has been
assumed in its proof. As we have noted, Steven assumes that the
force on a vertical surface of a given area at some level is equal to
the force on a horizontal surface of the same area at the same level.
It is because that latter force is independent of the lateral extent of
the water that the former is too. When Stevin continued the pas-
sage quoted immediately above with the words ‘Of these matters
5
we do not draw up any special propositions, in view of their
aforesaid clearness’ he gave the impression that his hydrostatics
had clearly explained phenomena associated with lock gates when
it fact it had done nothing of the kind.
The fact that Stevin inserted assumptions into his theory that
could be argued to be in need of explanation is connected with my
third point. Stevin’s theory did not provide clockwork explanations
of hydrostatic phenomena. He gave no precise, theoretical account
of how forces are transmitted from one part of a liquid to another
thereby bringing about hydrostatic effects. A number of Stevin’s
successors who built on his hydrostatics found this aspect of it
problematic. Isaac Beekmann and Rene Descartes, in 1619,
attempted to address what they saw as this deficiency in Stevin’s
theory by proposing corpuscular mechanisms, but without much
success (Descartes, 1964e76, Vol. 10, pp. 67e74). Decades later
Robert Boyle (2000, Vol. 5, pp. 207 and 236) was to complain that
Stevin had merely asserted that certain propositions in hydrostatics
are true without showing why they are true. By that time, Boyle was
able to draw on Pascal’s solution of the problem that we discuss
below.
4.5. The practice of hydrostatics and empirical support
In the Practice of Hydrostatics Stevin describes a number of
practical situations to which his theory of hydrostatics can be
applied. It is natural for a modern reader to interpret these as ex-
periments providing evidence for the theory, and, indeed, they
were appropriated and used in that way by Pascal. However, Stevin
made it quite clear that this was not how he interpreted the situ-
ation. The practical situations that he describes are intended by him
to be illustrations and applications of, not evidence for, his hydro-
statics. From Stevin’s point of view seeking support for his hydro-
statics in experiments would have been as wrong-headed as testing
Pythagoras’s theorem by measuring the sides of a material triangle.
The bulk of the Practice of Hydrostatics is concerned with some
of the counter-intuitive consequences of Proposition X, such as the
need for a heavy weight to support a much lighter body of water in
the situation depicted in my Fig. 1.4 When introducing his
description of the relevant ‘experiments’ Stevin made his views on
their purpose quite explicit.
We have proved mathematically in the 10th proposition of the
elements of hydrostatics, in the fifth corollary [that the force on
a horizontal surface depends only on its area and depth and is
otherwise independent of the volume of water]. But because
many people may consider this unnatural, we will, in addition to
the foregoing mathematical proof, describe five practical ex-
amples thereof, which anyone may test and see with his own
eyes (Boyle, 2000, p. 487).
One ‘application’ of his hydrostatics described by Stevin is worth
mentioning because it is so straightforward and simple to execute.
If one end of a hollow tube is blocked by a metal disc and the
arrangement is plunged disc-downwards into water with the upper
end of the tube open to the air, then the disc remains suspended in
spite of being denser than the water.
The stance taken by Stevin on the relationship between his El-
ements of Hydrostatics and practical applications is in line with the
metaphor he employed to introduce The Practice of Weighing as a
sequel to his statics.
Just as it would be useless to lay large and strong foundations
which can support a heavy edifice without ultimately wishing to
erect any building thereon, thus in the elements of the arts,
theory is lost labour when the end does not tend to practice. Just
as, in the natural order of things, the foundations precedes the
6 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10

building, thus theory precedes practice. This being so, it is 5. Pascal’s hydrostatics: introduction of the concept of
appropriate that, having described the Elements of the Art of pressure
Weighing, hereinbefore, I should follow this up with the Practice
of Weighing (Boyle, 2000, p. 293). Pascal’s theory of hydrostatics is based on a clear account of how
liquids, as such, transmit forces applied to them and thereby bring
The foundations of hydrostatics are provided by its ‘Elements’
about hydrostatic phenomena. In the following I summarise the
that are mathematically proven. Having been established they can
essentials of that theory staying close to Pascal’s own mode of
be put into practice in novel as well as familiar situations. To the
expression.
extent that my appraisal of Stevin’s hydrostatics is legitimate
In the first Chapter of his Treatise Pascal describes a number of
then, by expressing himself in this way, he is misrepresenting the
puzzling phenomena as a prologue to his theory which will explain
relationship between his theory and practice.5 On my view, Stevin
them and many more. The first five situations depicted in the top
was able to arrive at key theorems in his hydrostatics only by
row of Fig. 3, which is a reproduction of Pascal’s Plate 1, involve
feeding into it knowledge, like the horizontal force on a lock gate,
equal stoppers in the base of containers in which water of various
with which he was thoroughly acquainted as a practising
amounts reach the same height. Pascal (1937, p. 1) bluntly asserts
engineer.
that ‘experiment shows that it takes the same force to keep the

Fig. 3. Pascal’s Plate 1 in his Treatise on the equilibrium of liquids. From Pascal (1937, p. 21).
 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10
stoppers in, although there are very different amounts of water in
the several receptacles’. The discussion of Chapter 1 is extended to
include the situation in the sixth of Pascal’s figures, where the
column of water in the vertical tube on the left supports the much
greater weight bearing down via the broader vertical tube on the
right. These cases show that ‘a fine thread of water can balance a
heavy weight’ and Pascal sets himself the task of demonstrating
‘the cause of such multiplication of force’ (Pascal, 1937, p. 6).
The essentials of Pascal’s theory can be usefully articulated by
reference to what has since become known as the hydraulic press.
This is, in effect, what Pascal himself did in Chapter 2 of his Treatise.
Fig. 4 is my version of the seventh of the figures appearing in
Pascal’s Plate 1. The effect of W1 on the water beneath is described
by Pascal as follows:
Thus if a vessel filled with water has but one aperture, say one inch
in area, and a piston is placed on it under a one-pound weight,
that weight is exerted on every part of the vessel because of the
continuity and fluidity of the water. To ascertain how much of the
weight is borne by each part, the following rule holds good.
Each area equal to that of the aperture, that is, one square inch, is
pressed upon as if by a one-pound weight. (The weight of the
water is not taken into consideration here since only the weight
of the piston is being dealt with.) The one-pound weight presses
the piston at the aperture, and each part of the vessel is more or
less pressed in proportion to its area, whether that part be
opposite the aperture, or to one side of it, or far, or near; for the
continuity and fluidity of the water make all these circum-
stances equal and indifferent (Pascal, 1937, pp. 7e8.).6
This passage alone illustrates three key features of Pascal’s hydro-
statics. Firstly, we have a quantitative statement of the way in
which the force exerted by W1 is transmitted to other parts of the
container. It is transmitted as a force per unit area, one pound per
square inch in this case. Secondly, we have a clear indication of the
isotropic character of the force so transmitted, since the water
presses equally on any surface of a given area whatever its orien-
tation. Thirdly, Pascal makes it clear that it is a property of water as
such (which it shares with other liquids) that lies behind the
isotropic transmission of forces. They arise as a result of the ‘con-
tinuity and fluidity’ of water.
These basic ideas are elaborated and extended in Pascal’s short
thesis. When two weights are involved, as in the hydraulic press,
equilibrium will obtain when the weights are proportional to the
areas over which they act on the liquid. At equilibrium ‘the water is
equally pressed upon under the two pistons; for although one of
these is one hundred times as heavy as the other, it is, on the other
hand, in contact with an area one hundred times greater’ (Pascal,
1937, p. 8). The quantitative measure of the pressing of the water
is as a force per unit area.
Fig. 4. The hydraulic press. A schematic version of the seventh figure in Pascal’s Plate 1.
7
The isotropy of hydrostatic forces was reinforced in a number of
ways. For instance, when a column of mercury is supported in a
vertical tube immersed in water with its upper end open to the air,
the height of mercury supported is independent of whether the
lower end of the tube points straight down, is bent at right angles or
curves upwards (Pascal, 1937, pp. 12e13). A balloon filled with
water will uniformly swell or contract as it is raised or lowered in
water. ‘This is because the weight of the mass of water above
the balloon presses it on all sides towards the centre’ (Pascal, 1937,
p. 21).
Pascal’s considerations were not confined to situations where
the forces on the water originate from weights. The forces could, for
instance, arise from the action of men pressing down on the ap-
ertures, in which case ‘one man pressing on the smaller piston will
exert a force equal to that of one hundred men pressing on the
larger, and will exceed that of ninety-nine men doing the same’
(Pascal, 1937, p. 6). The weights bearing down on the water can just
as readily be replaced by volumes of water of equal weight, in
which case equilibrium will obtain when the heights of the col-
umns of liquid are the same. We have an explanation of why water
finds its own level and why the degree to which a liquid presses is
proportional to depth (Pascal, 1937, pp. 8e9 and p.12). Further, once
we have replaced the weights depicted in Fig. 3 by columns of
water, then, instead of weights pressing against water at the ap-
ertures we have water pressing against water. Pascal in effect gives
a clockwork explanation of how isotropic hydrostatic forces are
transmitted, by way of the pushes exerted by portions of water
against neighbouring portions or against a solid surface with which
they happen to be in contact.
The extent to which Pascal’s theory involves clockwork expla-
nations is very much to the fore in his discussion of the upthrust on
solids immersed in water.
We have seen that water presses upward bodies that it bears
upon from below, that it presses downward on those that it
bears up from above, and that it presses to one side those that it
bears upon from the opposite side. From this it can be readily
inferred that, when a body is wholly submerged, then, since the
water bears upon it above, below, and on every side, it strives to
push it up, down and to all sides; but as its head is the measure
of its force in all these efforts, there is no difficulty in deter-
mining which of them should overbear the rest. It is obvious at
once that since the water has the same height on all the lateral
faces, it will press upon them equally; — But as the water has a
greater head against the bottom than against the top, it will
obviously press the body more upward than downward; and
since the difference between these heights of water is the height
of the body itself, it will be readily understood that the water
presses it upward and not downward, and with a force equal to
the weight of a volume of water equal to that of the body (Pascal,
1937, p.16).
Here Pascal derives Archimedes’ principle by considerations of the
way in which weight acting downwards on a body is countered by
the results of water pressing against it. By adding the pressing of
the water to weight Pascal has, in this context, fulfilled the promise
made in the subtitle of his Treatises and identified the causes of the
phenomenon in question.7 In my terminology, he has supplied a
clockwork explanation of the phenomenon of floating and it is in
this way that he is able to remove qualms concerning the lack of
causal explanation in Stevin’s hydrostatics.
The identification of clockwork explanations in Pascal’s hydro-
statics can be taken a step further. In the second of his Treatises, On
the Weight of the Mass of the Air, Pascal (1937, pp. 70e71) contests
the claim, still common in the first half of seventeenth century, that
water does not weigh in water, claiming that he had already shown
8 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10
the contrary in The Equilibrium of Liquids. As a matter of fact, that
matter is not discussed directly in the earlier work. However, it is
quite easy to see how Pascal’s discussion of bodies immersed in
water can be extended to suit his purpose. A body of water
immersed in water will be pressed on all sides just like a solid body,
and, what is more, the upward force on the lower surface will
exceed the downward force on the upper surface because of the
difference in height. The resulting force upwards will be exactly
countered by the weight of the body of water. The pressing and the
weighing combine to yield a clockwork account of the situation,
and implies the way in which forces are transmitted through liquids
by way of the pushing of a portion of liquid against neighbouring
portions.
Once he had comprehended hydrostatics in terms of the
isotropic transmission of pushes through liquids as a force per unit
area, Pascal could compare devices such as the hydraulic press to
long-familiar machines such as the lever and pulley. All are alike
insofar as each represents ‘a mechanical machine for multiplying
force’. But there are differences that stem from the categorical
distinction between solids and liquids. The ‘clockwork’ mecha-
nisms responsible for the action of levers and pulleys rely on the
rigidity of levers and the limited extendibility of pulley strings.
These constraints arise from properties characteristic of solids that
were part of common sense and were taken for granted. In the case
of mechanisms involving liquids the precise nature of the con-
straints at work were not part of common sense and needed to be
articulated. That, in effect, is precisely what Pascal does with his
theorisation of what is implied by the ‘continuity and fluidity’ of
liquids.
The reliance of the functioning of hydraulic machines on this
characteristic feature of liquids is highlighted by Pascal by reference
to what happens if the water is turned to solid by freezing it. If the
water immediately under the weights in the hydraulic press is
frozen, leaving the water in the main vessel, via which they
communicate, unfrozen, then the functioning of the press is un-
impaired. On the other hand it does not function at all if that main
body of water is frozen.
From this it is most clearly evident that it is the liquidity of the
substance by which one aperture communicates with the other
that causes this multiplication of force. The fundamental reason
is, as we have said, that a vessel full of water is a mechanical
machine for multiplying force (Pascal, 1937, pp. 10e11).
Given the content and consequences of Pascal’s hydrostatics as I
have summarised it above, I feel justified in following Duhem
(1905) and Dijksterhuis (1970, p. 69) and crediting Pascal with
the introduction of a concept of pressure, even though he did not
use that term or its French equivalent.8
6. Hydrostatics and experiment
I noted above Pascal’s appeal to experiment on the very first
page of The Equilibrium of Liquids. ‘Experiment shows’ that differing
amounts of water can weigh equally on a base of given area how-
ever puzzling this may appear. Pascal (1937, p. 75) again, on the
final page of his two treatises, explicitly invoked experiment ‘as the
true master that one must follow in Physics’. The case for Pascal’s
hydrostatics rested on the way in which a simple conception of the
transmission of forces through liquids can naturally and without ad
hoc adjustment straightforwardly explain a range of hydrostatic
phenomena, many of them subject to precise investigation under
experimental conditions.
Insofar as it was articulated by reference to and supported by
experiment, Pascal’s approach to hydrostatics stood in contrast to
that of Stevin. The content of Stevin’s theory, as expressed in
‘Stevin’s law’, came close to accommodating the concept of pres-
sure insofar it contained a specification of the force exerted by a
liquid on a solid surface, whatever its orientation. However, Stevin
was not free to articulate this notion because the method that he
explicitly advocated required that his hydrostatics be derived from
postulates sufficiently unproblematic to be accepted at the outset.
Stevin was able to introduce novel content into his hydrostatics
only by covertly, and perhaps unwittingly, drawing on knowledge
of the behaviour of liquids with which he was familiar as an hy-
draulic engineer. What is more, as we have noted, Stevin’s
Euclidean method required that he treat as mere applications of a
theory proved by other means what Pascal would treat as experi-
mental evidence.
The above remarks concerning the role of experiment in Pascal’s
hydrostatics are in need of qualification. There are reasons to doubt
that Pascal actually performed many of the experiments to which
he appealed in support of his hydrostatics, many of them raised by
Robert Boyle (2000, Vol. 5, pp. 206 and 255). Boyle mocked Pascal
by observing that some of his experiments required experimenters
to carry out their observations whilst twenty foot under water!9
Added to this is the fact, referred to above, that versions of many
of the experiments appealed to by Pascal had already been
described in Stevin’s Practice of Hydrostatics.
Are we to conclude from observations of the kind cited above
that Pascal’s appeal to experiment was unwarranted rhetoric and
should we join Peter Dear (1990, p. 681) in concluding that Pascal’s
work was not ‘natural philosophy’ it was ‘mathematics’? I aim for a
nuanced response to the situation that captures a sense in which
Pascal’s emphasis on novel experimental support in his hydrostatics
was misleading but avoids the implication that his references to
such support were mere rhetorical additions to a theory that was
basically mathematical.
In the light of Boyle’s critique, it seems certain that Pascal did
not perform all of the experiments he referred to and may well have
performed very few of them. I am inclined to downplay the epis-
temological significance of this. The new hydrostatics involved both
theoretical and empirical innovations. The development of the new
practice was insensitive to whether one individual or several
contributed to those innovations. Pascal could draw on experi-
ments he found described by Stevin sure enough, but he integrated
them into a conceptual framework he devised himself drawing on
experimental developments and theoretical reflections of the pre-
vious decade. When Boyle did perform improved versions of ex-
periments of the kind invoked by Pascal, the outcomes were in
conformity with the latter’s strictures.
These observations notwithstanding, there is something not
quite right about Pascal’s rhetoric with respect to experiment as far
as his hydrostatics is concerned. The applications described by
Stevin in his Practice of Hydrostatics were accepted as facts and
augmented by the likes of Mersenne by the 1640s.10 If we take
those results and add to them long known facts about the behav-
iour of liquids, such as floating, water finding its own level and so
on, then a strong case could be made for Pascal’s theory involving
pressure by appeal to the straightforward and uncontrived way it
could explain all of those facts. Novel experimental results, such as
the functioning of the hydraulic press, did in a sense add further
support, but support that was hardly needed. It is plausible to re-
gard the hydraulic press as an application of a theory well
confirmed by the straightforward way in which it could explain a
range of phenomena that were well known by the middle of the
seventeenth century. That is why Pascal could confidently invoke
the detailed outcome of experiments he may not have performed.
On the other hand, the pneumatics that Pascal helped to develop
did require novel experimental findings to reveal relevant phe-
nomena and provide crucial evidence. Pascal’s rhetoric about the
 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10
importance of novel experiments in hydrostatics is a projection
of what was a perfectly reasonable view in pneumatics onto
hydrostatics.
7. Concluding remarks
The move beyond the science of weight to hydrostatics involved
opening up a terrain where common sense knowledge of the time
was incapable of yielding evident postulates capable of providing a
secure and adequate basis for the new science. The kind of
complexity involved in the mutual relationship between theoret-
ical reflection and experiment echoes that involved in Galileo’s
mechanics, a topic that has been given much attention in the
literature. Alexander Koyré famously challenged the idea that
Galileo’s mechanics was established by generalising the results of
independently established experimental results. Noting that a
number of key experiments exploited by Galileo were thought
experiments that he could not have performed, Koyré (1968, p. 13)
concluded that “it was thought, pure unadulterated thought, and
not experience or sense perception — that gives the basis for the
‘new science’ of Galileo Galilei”. More subtle accounts of the rela-
tionship between theory and experiment in Galileo’s work that
moved beyond the dichotomy between empiricism and rationalism
that Koyré presupposed have since been forthcoming, exemplified
in Clavelin (1974) and Gaukroger (1978). Damerow, Freudenthal,
Mclaughlin, and Renn (2004) is a more recent, detailed and
discerning account of the emergence of the concept of motion
involved in classical physics as a result of attempts by Galileo and
his contemporaries to conceptualise long known as well as novel
experimentally produced phenomena involving motion. The com-
plex interchange of work by head and hand identified in that work
shares the qualitative features of what was involved in the emer-
gence of the concept of pressure.
The extent to which the notion of pressure was by no means
obvious and needed to be constructed can be illustrated by refer-
ence to the inadequacy of the hydrostatics of Galileo and Descartes
in this respect. Galileo based his hydrostatics on the principle that if
a system is displaced about an equilibrium position, then the ve-
locities of the displaced parts are inversely proportional to their
weights, a principle that Galileo had employed in the context of the
balance and which he attributed to Aristotle.11 If a floating body is
subject to a vertical displacement then its velocity of displacement
times its weight is equal to the velocity of the liquid displaced times
its weight. Galileo dealt with the hydrostatic paradox in a similar
way. Appeal to this principle enabled Galileo to show that specifi-
cations of equilibrium conditions were the correct ones once they
were known. However, since his considerations involved only
vertical displacements he was ill-equipped to accommodate the
isotropic character of hydrostatic forces and, what is more, he did
not provide clockwork explanations of hydrostatic phenomena.12
Descartes’ attempts to provide causal explanations for hydro-
static phenomena involved appeal to strings of corpuscles or par-
ticles of matter/extension each one pressing against and pressed by
its neighbours. This was true of his early attempt to explain some of
Stevin’s claims in corpuscular terms, persisted in his treatment of
wine pressing on a vat in Chapter 3 of Le Monde and remained the
case into his mature Principles of Philosophy, where hydrostatics
along with material phenomena generally were to be reduced to
pushes between particles of matter/extension induced by their
tendency to move away from the axes or centres of the vortices in
whose motion they participated. This approach prevented Des-
cartes from grasping the isotropy of hydrostatic forces and ensured
that he made no significant moves towards the concept of pres-
sure.13 He failed to provide adequate clockwork explanations of
hydrostatic phenomena that met his aspirations.
9
It was Pascal who first fashioned a concept of pressure in his
Equilibrium of Liquids. That Treatise involved a style of reasoning
that stood in contrast to Stevin’s that was modelled on Euclidean
geometry. There are two senses in which the new style could be
appropriately described as ‘experimental’. Experiment was
involved in the interrogation and exploration of hydrostatic phe-
nomena that, from Stevin to Pascal, took familiarity with the phe-
nomena beyond the range of common sense and which paved the
way for conceptual innovation. It was also experiment that pro-
vided some of the evidence on which the new hydrostatics was to
be based and against which it could be tested.
The formation of a notion of pressure up to the demands of
hydrostatics was a theoretical accomplishment as well as an
empirical and experimental one in spite of the fact that the notion,
once formed, was as close to what could be observed and measured
as the concept of weight. The theoretical conjectures that culmi-
nated in claims made using the concept of pressure should not be
identified with speculations about the ultimate structure of matter
in terms of interacting corpuscles characterised solely in terms of
their size, shape and motion that seventeenth-century ‘mechanical’
philosophers were concerned to defend as an alternative to scho-
lastic Aristotelianism.14 Such considerations have no useful place in
the story I have told above. Pascal’s hydrostatics was experimental
insofar as experiment was involved in the fashioning of the new
concept of pressure and in providing evidence for claims made with
its aid, it was theoretical insofar as the construction of the notion of
pressure involved intellectual labour and it was mechanical insofar
as it involved clockwork mechanisms comprehended via the notion
of pressure.
Acknowledgements
My research was supported by a Discovery Grant from the
Australian Research Council, Grant number, DP110102471, and
greatly facilitated by the hospitality of the Max Planck Institute for
the History of Science, to the research facilities of which I had access
for a total of five months. I also benefited from visits to the Phi-
losophy Departments at the University of Durham (two months)
and the University of Bristol (one month). The feedback from
seminars delivered and workshops participated in at those in-
stitutions have been invaluable. Antoni Malet, Matthias Schemmel,
John Schuster and Wolfgang Lefèvre read early drafts of this paper
and I thank them for their time, their encouragement and their
critical and constructive comments. I thank Michael Hanaghan for
help with texts in Latin.
References
Boyle, R. (2000). In M. Hunter, & E. B. Davis (Eds.), The works of Robert Boyle. London:
Pickering and Chatto (15 Vols.)
Chalmers, A. (2012). Intermediate causes and explanations: The key to under-
standing the scientific revolution. Studies in History and Philosophy of Science,
43, 551-562. http://dx.doi.org/10.1016/shpsa.012.06.005.
Clavelin, M. (1974). The natural philosophy of Galileo: Essay on the origins and for-
mation of classical mechanics. Cambridge, Massachusetts: MIT Press.
Damerow, P., Freudenthal, G., Mclaughlin, P., & Renn, J. (2004). Exploring the limits of
preclassical mechanics: A study of conceptual development in early modern sci-
ence. New York: Springer.
Dear, P. (1990). Miracles, experiments, and the ordinary course of nature. Isis, 81,
663-693.
Descartes, R. (1964-76). In C. P. Tannery (Ed.), Oeuvres de Descartes (2nd ed.). Paris:
Vrin (11 Vols)
Dijksterhuis, E. J. (Ed.). (1955). The principal works of Simon Stevin (Vol. 1)Amster-
dam: Swets & Zeitlinger.
Dijksterhuis, E. J. (1970). Simon Stevin: Science in the Netherlands around 1600. The
Hague: Martinus Nijhoff.
Drake, S. (1981). Cause, experiment and science. Chicago: University of Chicago Press.
Duhem, P. (1905). Le principe de Pascal. Revue general des sciences pures et appliques,
16, 599-610.
10 A.F. Chalmers / Studies in History and Philosophy of Science 51 (2015) 1e10
Gaukroger, S. (1978). Explanatory structures: Concepts of explanation in early physics
and philosophy. Hassocks, Sussex: Harvester Press.
Gaukroger, S., & Schuster, J. (2002). The hydrostatic paradox and the origins of
Cartesian dynamics. Studies in History and Philosophy of Science, 33, 535-572.
Koyré, A. (1968). Metaphysics and measurement: Essays in the scientific revolution.
London: Chapman and Hall.
Mersenne, M. (1644). Cogitata physico-mathematica. Paris: Antonii Bertier.
Palmieri, P. (2005). The cognitive development of Galileo’s theory of buoyancy.
Archive for History of Exact Sciences., 32, 109-129. http://dx.doi.org/10.107/
s00407-004-0089-2.
Pascal, B. (1937). The physical treatises of Pascal: The equilibrium of liquids and the
weight of the mass of the air (I. H. B. and A. G. H. Spiers, Trans.). New York:
Columbia University Press
Pascal, B. (1970). In J. Mesnard (Ed.), Blaise Pascal: Oeuvres Complète (Vol. 2)Paris:
Desclée de Brouwer.
Schuster, J. (2013). Descartes-Agonistes: Physico-mathematics, method and corpus-
cular mechanism, 1618e33. Dordrecht: Springer. http://dx.doi.org/10.1007/978-
94-007-4746-3.
Shapiro, A. (1974). Light, pressure and rectilinear propagation: Descartes’ celestial
optics and Newton’s hydrostatics. Studies in History and Philosophy of Science, 5,
239-296.
Endnotes
1
My references to Stevin’s works are to the English translations in E. J.
Dijksterhuis (1955) which are accompanied by the original Dutch versions. In the
main I have used the English translation of Pascal’s Treatise in Pascal (1937). Where
I have needed to cite the original French I have used the version appearing in Pascal
(1970).
2
Such phenomena came to be referred to as ‘hydrostatic paradoxes’ after Boyle
introduced that terminology in 1666.
3
Here I exploit Stevin’s own articulation of what I have referred to as the
Euclidean ideal, illustrated in the passages from The Art of Weighing cited in Section
3.
4
Stevin’s version of this diagram appears in Dijksterhuis (1955, facing p. 493).
5
The quotation is taken from a dedication of Stevin’s Practice of Weighing to the
burgomasters and rulers of the city of Nurenburg and so should be interpreted
bearing in mind his intent to present his book as serving their interests.
6
I have altered the English translation to avoid using the noun ‘pressure’,
thereby avoiding the danger of mistakenly attributing to Pascal what is entailed by
the modern, technical, usage of that expression. The French version, reads ‘de sorte
que si un vaisseau plein d’eau n’a qu’une seule ouverture, large d’un pouce, par
example, où l’on mette un piston chargé d’un poids d’une livre, ce poids fait effort
contre toutes les parties du vaisseau généralement, à cause de la coninuité et de la
fluidité de l’eau; mais, pour determiner combine chaque partie souffre, en voici la
règle. Chaque partie large d’un pouce, comme l’ouverture, souffre autant que si elle
était poussée par le poids d’une livre (sans compter le poids de l’eau dont je ne
parle pas ici, car je ne parle que du poids du piston) parce que le poids d’une livre
presse le piston qui est à l’ouvrture, et chaque portion du vaisseau plus ou moins
grande souffre précisément plus ou moins à proportion de sa grandeur, soit que
cette portion soit vis-à-vis de l’ouverture ou à côté, loin ou près; car la continuité et
la fluidité de l’eau rend toutes ces choses-là égales et indiffèrentes.’ (Pascal, 1970, p.
1046.)
7
The subtitle of Pascal’s Treatises reads ‘Containing the explanation of the causes
of various effects of nature which had not been known hitherto, and in particular of
those which had been ascribed to the abhorrence of a vacuum’ (Pascal, 1937,
p. xxix).
8
Pascal fell short of an explicit acknowledgement of the fact that within the body
of a liquid pressure acts equally in all directions. As Shapiro (1974, p. 277) has
documented, it was left to Newton to do that.
9
Witness the gentleman in the final figure of Pascal’s Plate I.
10
In his Cogitata Physico-mathematica Mersenne included a broad survey of de-
velopments in hydrostatics from Stevin onwards, embellished by astute observa-
tions of his own. For his summary, and acceptance, of Stevin’s contributions see
Mersenne (1644, pp. 217e219).
11
It has since been realised that the Mechanical Problems to which Galileo referred
was not by Aristotle himself but one of his followers.
12
Galileo’s hydrostatics appears in his Bodies that stay atop of water or move in it,
published in 1612. An English translation is in Drake (1981). See Palmieri (2005) for
a recent appraisal of Galileo’s hydrostatics.
13
See Descartes (1964e76, Vol. X, pp. 67e74) for his early manuscript on Stevin’s
hydrostatics. It is discussed in Gaukroger and Schuster (2002). Descartes’ hydro-
statics is discussed, and criticised, in Shapiro (1974, pp. 260e266) and Schuster
(2013, pp. 114e121 and 509e514).
14
In Chalmers (2012) I exploit Boyle’s distinction between intermediate and ul-
timate causes to help distinguish between the attempts to construct a mechanical
account of the ultimate structure of the world and attempts to explain phenomena
mechanically in a more common sense that includes the appeal to clockwork
mechanisms.