Modern Science and Refutation of the Paradoxes of Zeno

Author(s): Adolf Grunbaum

Source: The Scientific Monthly , Nov., 1955, Vol. 81, No. 5 (Nov., 1955), pp. 234-239
Published by: American Association for the Advancement of Science

Stable URL: https://www.jstor.org/stable/22052

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

American Association for the Advancement of Science is collaborating with JSTOR to digitize,
preserve and extend access to The Scientific Monthly

This content downloaded from

90.166.56.191 on Tue, 11 Mar 2025 09:19:29 UTC
All use subject to https://about.jstor.org/terms
 Modern Science and Refutation
of the Paradoxes of Zeno
ADOLF GRUNBAUM

Dr. Griinbaum is professor of philosophy at Lehigh University, Bethlehem,

Pennsylvania. He has contributed two other articles to The Scientific Monthly:
"Science and ideology," July 1954, and "Operationism and relativity," Oc-
tober 1954.

A BOUT 2500 years ago, a Greek named tradictory to maintain that a positive time interval
Zeno of Elea confounded his contempor- can consist of instants of zero duration. Among,
aries by a series of startling arguments. those who have taken up their cudgels on this
These were designed to show that the science of issue, we find such diverse thinkers as Kant (3),
geometry is beset by a para4ox and that any at- P. du Bois-Reymond (4), William James (5), and
tempt to provide a mathematical description of P. W. Bridgman (6). Yet the very conception that
motion becomes ensnared in contradictions. So Zeno and these writers would proscribe has become
seminal was the scientific challenge bequeathed to commonplace in contemporary science through the
posterity by Zeno's polemic that the contemporary work of the German mathematician Georg Cantor
philosopher Bertrand Russell paid tribute to him (7, p. 275). It therefore behooves us to come to
(1), saying: "Zeno's arguments, in some form, grips with this charge of absurdity.
have afforded grounds for almost all the theories Consider the computation by means of which
of space and time and infinity which have been Zeno deduced his paradoxical result. In order to
constructed from his day to our own." add the lengths of all the points in an interval of
Authorities disagree on the identity of the ad- 2 centimeters, we must first form a precise idea of
versaries at which our searching Greek aimed his how many points compose such an interval. We
intellectual broadside. Disregarding entirely the note immediately that the number of points here is
question of historical authenticity, we shall con- not finite. But of what avail is the trite observation
sider a version of Zeno's paradoxes directly rele- that it is infinite? To no avail, unless we succeed
vant to modern science. To my knowledge, the in giving a mathematically articulate characteriza-
arguments that I shall offer in refutation of Zeno tion of the particular kind of infinite collection
have not been given by previous writers. It will be with which we are confronted. It was Cantor's
best to examine (i) the geometric paradox, which achievement to provide precisely this characteriza-
impugns the consistency of the contemporary math- tion. And since it will give us the means by which
ematician's conception of the relationship between to disprove the Zenonian allegation against ge-
a line and its points, and (ii) the paradoxes of ometry, we briefly pause to give a statement of its
motion by which Zeno attempted to demonstrate meaning.
the impossibility of motion. Suppose that a very intelligent child who does
not, however, know the names of numbers exceed-
Geometric Paradox ing 10 is confronted with two bags of pennies, each
In the geometric paradox, our philosopher asserts containing more than 10 pennies, and is then asked
that it is self-contradictory to claim that a line to determine which of the two bags contains more
segment consists of points, each having zero length pennies. By virtue of his limited knowledge of
(2). For if a line segment of, say, 2 centimeters numbers, the child cannot make a separate count
actually does consist of points, then the total length of the contents of either bag. But his intelligence
of that segment should be computable by adding will enable him to make the required determina-
the individual lengths of its constituent points. But tion nonetheless: he will pair off each coin in bag
instead of yielding the required value of 2 centi- A with a specific coin in the other bag B. If he
meters, this computation unavoidably yields the exhausts the coins in bag B and still has coins left
paradoxical result of 0 centimeters, since a summa- over in bag A, then he will corfclude that there are
tion of zeros can issue in nothing other than zero. more pennies in bag A than in bag B. On the other
By the same token, Zeno argues that it is self-con-
hand, if the supply in B outlasts the supply in A,

234 THE SCIENTIFIC MONTHLY

This content downloaded from

90.166.56.191 on Tue, 11 Mar 2025 09:19:29 UTC
All use subject to https://about.jstor.org/terms
he will know that the B collection is richer than contain the same super-denumerable infinity of
the one in A. And if he comes out even, he will points. Therefore, if the coxncept of super-denu-
know, without having been able to make separate merable infinity is itself free from contradictions,
counts, that bag A contains the same number of then Zeno's charge of inconsistency is false, both
pennies as bag B. By the same kind of reasoning, in regard to the decomposition of a line into math-
we know that, in a monogamous society, there are ematical points and with respect to the resolution
just as many husbands as wives, although we have of a time interval into instants of no duration (8).
not taken a census. Proponents of Zeno's view might still argue that
Cantor saw that this method of one-to-one pair- this arithmetical rebuttal is unconvincing on purely
ing lends itself to the solution of the problem of geometric grounds, maintaining that if extension
comparing two infinite collections as to cardinality: (space) is to be composed of elements, these must
two such collections are equinumerous, if there is themselves be extended. Specifically, geometers
at least one way of pairing off their members so like Veronese objected (9) to Cantor that in the
as to secure for each member in either collection array of points on the line, their extensions are all,
exactly one partner in the other. In the event that as it were, "summed geometrically" before us. And
no such pairing can be achieved, the collection from this geometric perspective, it is not cogent, in
that is always left with some unpaired objects is their view, to suppose that even a super-denumer-
the larger of the two. Now, Cantor was able to able infinity of unextended points would be able
show (7, pp. 278-280) that the infinity of points to sustain a distance, especially since the Cantorean
in a unit line segment is larger in precisely this theory can claim arithmetical consistency here only
sense than the infinity of positive integers. Having because of the obscurities that obligingly surround
called the latter infinity "denumerable," he then the meaning of the arithmetic "sum" of a super-
called the former "super-denumerable." denumerable infinity of numbers.
In the context of modern mathematics, Zeno Is this objection to Cantor conclusive? I think not.
is thus defying us to obtain a result differing from Whence does it derive its plausibility? It would
zero upon adding all the lengths of the super-de- seem that it achieves persuasiveness via a tacit ap-
numerable infinity of points that compose a unit peal to a pictorial representation of the points of
segment. This means that we are being asked to mathematical physics in which they are arrayed
add as many zeros. To Zeno's mind, it was axio- in the consecutive manner of beads on a string to
matic that the result of any addition of zeros would form a line. But the properties that any such rep-
be zero, regardless of the cardinality of the set of resentation imaginatively attributes to points are
zeros to be added. But he could not anticipate not even allowed, let alone prescribed, by the for-
that the addition of a super-denumerable infinity mal postulates of geometric theory. The spurious-
of numbers, be they zero or positive, presents a ness of the difficulties adduced against the Can-
problem altogether different from adding either a torean conception of the line becomes apparent
finite sequence of numbers such as 3, 4, 7 or a upon noting that not only the cardinality of its
denumerable infinity of numbers such as 1, 1/2, constituent points altogether eludes pictorialization
1/4, 1/8, 1/16, 1/32, 1/64, .... Although arithme- but also their dense ordering: between any two
tic has evolved a definition of the "sum" of a de- points, there is an infinitude of others. Thus, in
numerable infinity of numbers by a consistent complete contrast to the discrete order of the beads
generalization of the concept of finitary sum on the on a string, no point is immediately adjacent to
basis of the limit concept, this definition is utterly any other.
useless and irrelevant when the addition of a super- These considerations show that from a genuinely
denumerable infinity of numbers is called for. geometric point of view, a physical interpretation
Thus, the addition of the lengths of all the points of the formal postulates of geometry cannot be
composing a unit segment is a meaningless (un- obtained by the inevitably misleading pictorializa-
defined) operation in analytic geometry and hence tion of individual points of the theory. Instead, we
cannot be used to compute a value for the total can provide a physical interpretation quite un-
length of the segment. In particular, the deduction encumbdred by the intrusion of the irrelevancies
of Zeno's paradoxical result zero for the length of of visual space, if we associate not the term point
a unit segment is precluded by this arithmetical but the term linear continuum of points of our
fact. And what holds for the unit segment also theory with an appropriate body in nature. By a
holds for a segment of 2 centimeters and for any point of this body we then mean nothing more or
other, for, mirabile dictu, Cantor demonstrated less than an element of it possessing the formal
that no matter what their length, all line segments properties prescribed for points by the postulates

November 1955 235

This content downloaded from

90.166.56.191 on Tue, 11 Mar 2025 09:19:29 UTC
All use subject to https://about.jstor.org/terms
of geometry. And, on this interpretation, the of an infinite series of subsidiary time intervals
ground is then cut from under the geometric parti T/2n (n =. . ., 3, 2, 1), which has no first ter
pris against Cantor by the modern legatees of Zeno. because the denseness postulate entails infinite
Apart from the metrical consistency of the divisibility, and (ii) how can a temporal process
modern conception of an interval as an aggregate be completed in a finite time, if its completion
of unextended points, mathematicians have suc- requires the elapsing of an endless progression of
ceeded during the current century in proving, by temporal subintervals T/2n (n- 1, 2, 3, . . .)
means of the topological theory of dimension, that which there is no last term as a consequence of
there is also no contradiction in regarding the one- the denseness postulate?
dimensional line as consisting of zero-dimensional The more familiar versions of these two queries
points (10). This proof is not redundant with the are the "Dichotomy" paradox and the paradox of
foregoing analysis, since it turns out that the con- "Achilles and the Tortoise." In the former, whose
cepts of length and one-dimensionality are not at argument was endorsed by A. N. Whitehead and
all the same. P. Weiss (12), Zeno contends that the mathemati-
In addition to refuting Zeno's geometric para- cal theory of motion entails the impossibility of the
dox, our analysis yields the following result: unless very process that it purports to describe, for if
substantial modifications are made simultaneously there is indeed no point next to the starting point
throughout the body of analytic geometry, the on the path of a runner that he can occupy im-
proposal of some writers that we replace the mediately after leaving the starting point, then an
Cantorean conception of the line as continuous by infinity of points and, hence, intervals inevitably
the postulate that it consists of only a denumerable, always keep interposing themselves between the
discontinuous infinity of points must be rejected on starting point and any other point to which the
logical grounds of inconsistency alone (11). For runner would move. And thus the motion is nipped
the length or measure of a denumerable point set in the bud, as it were. On the other hand, even if,
is zero, as can be seen upon denumerating any per impossibile, the motion had begun, the para-
such set and then applying the familiar arithmetic dox of Achilles shows that its completion in a
definition of the sum of a denumerable sequence finite time is quite unachievable. Having granted
to the sequence of zeros representing the lengths a head-start d to the slow tortoise, which moves
of its members. with velocity v, Achilles can never catch up with
the turtle, despite his greater speed V: the fleet-
footed warrior will not overtake the reptile after
Paradoxes of Motion
a time d (7V -v), because there is no last term
Among the four paradoxes with which Zeno in the series of time intervals
sought to discredit the possibility of physical mo-
d d - d
tion, only two are of relevance to contemporary -+V + v2 + * *3
mathematical physics. Their pertinence derives
from that discipline's affirmation that the time a series whose terms represent the successive times
variable ranges over the real numbers just as the required by Achilles to traverse the successive dis-
space variable does. More particularly, it is the tances separating him from the tortoise, beginning
denseness of the ordering of these numbers, when with the head-start d of the tortoise.
they are arranged according to magnitude, that According to a view that is as widespread as it is
provides the point of application for Zeno's po- erroneous, Zeno's argument is no more than a
lemic. The claim that for any point on the path mathematical anachronism. We are told that if he
of a moving object, there is no next point, any- had only known, as we do today, that the arith-
more than there is an immediately following or metic sum of an infinite convergent series is finite
preceding instant for any instant during the mo- rather than infinite, then he would have recog-
tion, enables Zeno to ask incisively: In what sense nized that he had merely posed a pseudo-problem.
can the events composing the motion be signifi- More particularly, the sum formula for a geo-
cantly said to succeed one another temporally, if metric series would have convinced him at once
they succeed one another densely rather than in that the sum of the series considered in the Dichot-
the consecutive manner of a discrete sequence? omy is actually T and that the series of times
This question takes the form of asking (i) how can needed by Achilles to come abreast of the tortoise
a temporal process even begin, if, in order to does add up to the finite value d (V - v). This
survive the lapse of a positive time interval T, a retort is based on the fact that in modern mathe-
body must first have endured through the passage rnftics. the arithmetic "sum" of an infinite con-

236 THE SCIENTIFIC MON'THLY

This content downloaded from

90.166.56.191 on Tue, 11 Mar 2025 09:19:29 UTC
All use subject to https://about.jstor.org/terms
vergent series is defined as the limit of the sequence sion merely in retrospect, on the one hand, and, on
of its partial sums. An illustration will serve to the other, an as yet unbegun or uncompleted tem-
introduce my contention that this rebuttal falla- poral process, whose very inception or consum-
ciously attempts to settle a question of physical fact mation is first at issue, as we saw in the case of the
by invoking a definition. recital of the set of positive integers.
Suppose that there were no limitations on the What is required in order to refute Zeno's ob-
rapidity of human speech or thought, so that a jections to the mathematical theory of motion is a
man could recite the numbers 1, 2, 3, 4, . . . one proof that neither the denseness of the ordering of
by one, taking 1 minute to utter the first, Y2 min- the constituent events of the motion nor such fea-
ute for the second, '4 minute for the third, and tures of this process as are entailed by this dense-
continuing to require for each successive vocal act ness property constitute obstacles to its inception
only half the time of its predecessor. The times and consummation.
taken to recite the infinite series of positive integers To lay the groundwork for this proof, we ask:
would therefore form the corresponding infinite What is the basis for the view that the very mean-
series 1, 1 /2, 1/4, 1/8, 1/16, . . . . Remembering ing of temporal succession involves that events
that the series 1, 2, 3, 4, . . . has no last term, we follow upon one another seriatim, like the consecu-
now ask: Will it take our hypothetical man a finite tive beats of the heart, and not densely? The
or infinite amount of time to recite the totality of answer can be none other than that this feeling
positive integers? Clearly, this question is one of derives from a tacit appeal to the properties of the
physical fact. And no such question can ever be time flow experienced intuitively in human con-
settled by having recourse to an arbitrary defini- sciousness. Since each act of thought takes a posi-
tion of what we shall mean by the "sum" of an tive amount of time rather than a mere instant of
infinite convergent series. Mathematically, it is zero duration, it is inevitable that upon analyzing
entirely a matter of stipulation, as H. Hankel has the stream of consciousness into a succession ol
explained (13), whether we define that sum to be constituent moments or "nows," these elements
finite in the customary manner or infinite. In any are experienced as occurring in a discrete sequence.
physical situation, it is therefore an open, empiri- No wonder therefore that on such an intuitively
cal question which one of these two logically pos- grounded meaning of temporal succession, there is
sible definitions is relevant. an ever-present feeling that if physical events are
It follows that the adoption of the first of these to succeed one another in time, their order of
two definitions by the arithmetician cannot con- occurrence must also be discrete, if it is to be a
tain the answer to the telling question that Zeno temporal order at all. It follows that refutation of
raised when he asked, in effect: How can the de- Zeno will be at hand, if the psychological criterion
scription of motion resulting from the adoption of of temporal sequence can be supplanted by a
the customary limit definition of the "sum" of an strictly physical criterion whose definition of "event
infinite series in arithmetic be shown to be the B is later than event A" does not entail a discrete
physically true one in the face of the physical diffi- temporal order but allows a dense order instead!
culties exhibited by the Dichotomy and Achilles Fortunately, the definition of time order and
paradoxes? To be sure, if it can be shown, in- direction made possible by a careful interpretation
dependently of the meaning that arithmetical of the second law of thermodynamics provides a
theory decrees for the term sum, that there are no criterion with precisely these required properties.
physical reasons after all that would preclude the Subject to a qualification to be mentioned pres-
inception and completion of the motion in a finite ently, this definition can be stated as follows: Of
time, then, of course, the familiar summation pro- two entropy states of a closed nonequilibrium sys-
cedure used in arithmetic will tell us the length of tem, the state of higher entropy will be called the
the required time interval. It will not do to object temporally later state. And we see now that the
at this point by saying that a mere glance at a entropies of a closed system, whose values are given
unit line segment establishes that the customary by real numbers, can entirely intelligibly charac-
limit definition of "sum" is the physically relevant terize a set of physical states that form a linear
one, as is seen upon subdividing the unit segment Cantorean continuum and are thereby densely
into an infinite sequence of diminishing subinter- ordered with respect to the relation "later than."
vals I/2, I4, I8, . . . , all of which must be con- The qualification to be mentioned is made neces-
tained in the original unit segment. This objection sary by the fact that the so-called "periodicity" and
offers a false analogy between a static whole al- "reversibility" objections of statistical mechanics
ready present in its entirety and subjected to divi- have shown that the entropy of a single closed svs-

November 1955 237

This content downloaded from

90.166.56.191 on Tue, 11 Mar 2025 09:19:29 UTC
All use subject to https://about.jstor.org/terms
tem does not increase, monotonically with time. purely mathematical decompositions of the motion
A viable definition of time order on the basis of into a discrete denumerable infinity of subintervals.
the second law of thermodynamics must therefore Neither can he create difficulties by pointing out
utilize a more complicated definition that makes that if we consider the series of individual point-
reference to the statistical behavior of a class of events that divide one subinterval from the next
closed systems rather than raerely to a single such in the Achilles paradox, instead of considerirng
system. For the numerous relevant detailsy I must these subintervals themselves, then indeed we are
refer the reader to other publications (14), where confronted with an unending progression of physi-
he will also find my reply to the operational cri- cally distinct events, all of which must have elapsed
tique of the entropic definition of time order by the time the motion is completed. For by Zeno's
offered by P. W. Bridgman (15). own standards, as enunciated in connection with
Our analysis has shown that we are absolved the geometric paradox of extension, no less than
from the necessity of answering "how" a sUccession
by the results of modern measure theory, the total
of events can occur by exhibiting a discrete se- time required for the occurrence of that discrete
quence of occurrence. Upon freeing ourselves from denumerable sequence of point-events is zero. Ac-
the limitations of the psychological criterion of cordingly, this sequence, although unending, can
time order by means of the constructive elabora- hardly preclude the occurrence of the terminal
tion of an alternative, autonomous physical cri- event of the motion after a finite time.
terion, it becomes clear that the dense temporal
ordering of the constituent point-events of a mo- Quantum Theory
tion is no obstacle whatever to either its inception
or its completion in a finite time. And thus it is What is the essential bearing of the quantum
seen to be entirely unwarranted to ask "how" the theory on the argument offered here? The hypo-
motion can occur despite the dense temporal thesis of the atomicity of change has taken the
order, or what the runner does immediately after form of postulating that there are minimal proc-
leaving his point of departure. esses in nature in the sense that no changes occur
Given that it is established now that the entire and remain below certain spatial and temporal
dense set of events constituting the motion can minima. The proposed minimal displacement or
occur, we can, of course, subdivide it mathemati- length, the "hodon," is of the order of 1013 centi-
cally into various kinds of spatial and temporal meter, the diameter of an electron. And the mini-
subintervals. Thus, the distance traversed can be mal duration, or "chronon," is of the order of 10-23
subdivided into an arbitrarily large finite number second, the time required by electromagnetic dis-
of equal subintervals. By the same token, the dense turbances, the fastest in nature, to traverse a hodon.
ordering allows us to subdivide the motion mathe- Thus, space, time, and change are quantized in
matically into a regression of decreasing time or the same contingent physical sense in which masses
space subintervals, having no first subinterval, as in are quantized by the existence of an elementary
the case considered in the Dichotomy. Or, we can particle of minimum mass. Being extensive magni-
mathematically effect a subdivision into a progres- tudes, whose values are given by real numbers,
sion of decreasing subintervals, which has no last "atoms" of space or time presuppose logically all
subinterval, as is illustrated by the mathematical the constituent parts of which they can be regarded
decomposition of the motion of Achilles, But it to be the sum. In each of the finite number of
would be a grievous error to infer that the indi- hodons that the runner traverses, he goes through
vidual terms of these infinite series denote physi- a continuum and hence a dense infinity of points
cally distinct steps into which the motion has been and instants! The "atomicity" of physical processes
chopped up, and which the runner executes stac- affirmed by quantum theory is therefore anchored
cato rather than legato, as musicians would say. in a mathematical continuum. And the results of
For no infinite set of distinct, consecutively ordered that theory do not render our refutation of Zeno
operations would constitute a physical process that either superfluous or invalid.
is temporally continuous in the mathematical sense, In regard to theories of space or time quantiza-
as the runner's motion is assumed to be. And any tion that are not grounded in a mathematical con-
such discontinuous physical process would indeed tinuum but are thoroughgoingly atomistic, it will
defy completion in a finite time: the execution of suffice to cite the following perceptive comment by
the geometric progression of physically distinct H. Weyl (17): "So far, the atomistic theory of
acts of reciting the positive integers in the manner space has always remained mere speculation and
described earlier would take forever (16). has never achieved sufficient contact with reality.
Zeno can therefore not embarrass us by his How should one understand the metric relations

238 THE SCIENTIFIC MONTHLY

This content downloaded from

90.166.56.191 on Tue, 11 Mar 2025 09:19:29 UTC
All use subject to https://about.jstor.org/terms
in space on the basis of this idea? If a square is 11. For details on such proposals, see sec. 7, pp. 302-304
in reference 8. I am indebted to the National Science
built up of miniature tiles, then there are as many
Foundation and to the Lehigh Institute of Research
tiles along the diagonal as there are along the side; for support of research in the philosophy of science.
thus the diagonal should be equal in length to the 12. A. N. Whitehead, Process and Reality -(Macmillan,
side." New York, 1929), pp. 53, 96, 105-107, and Science
and the Modern World (Macmillan, New York,
References and Notes 1925), p. 186; P. Weiss, Reality (Princeton Univ.
Press, Princeton, N.J., 1939), chap. 6 and 7. See
1. B. Russell, Our Knowledge of the External World also Galileo Galilei, Dialogues Concerning Two New
(Allen and Unwin, London, 1914), p. 183. Sciences, H. Crew and A. de Salvio, Trans. (Mac-
2. See H. Hasse and H. Scholz, Die Grundlagenkrisis millan, New York, 1914), pp. 162-164.
der griechischen Mathematik (Pan, Charlottenburg, 13. H. Hankel, Theorie der komplexen Zahlensysteme
1928), p. 11.
(Voss, Leipzig, 1867).
3. I. Kant, Critique of Pure Reason, N. K. Smith, 14. E. Schr6dinger, "Irreversibility," Proc. Royal Irish
Trans. (Macmillan, London, 1929), pp. 203-204.
Acad. 53 (sec. A), 189 (1950); H. Reichenbach,
4. P. du Bois-Reymond, Die allgemeine Funktionen-
"Les fondements logiques de la mecanique des
theorie (Laupp, Tiubingen, 1882). Pt. I, p. 66.
quanta," Ann. inst. Poincare' 13, 140-158 (1953);
5. W. James, Some Problems of Philosophy (Long-
A. Griinbaum, "Carnap's views on the foundations
mans, Green, London, 1911), pp. 185-186.
of geometry," in P. A. Schilpp, Ed., The Philosophy
6. P. W. Bridgman, "Some implications of recent points
of Rudolf Carnap, Library of Living Philosophers,
of view in physics," Rev. Intern. Phil. 3, 490 (1949).
vol. X (Tudor, New York), in press; A. Griinbaum,
7. G. Cantor, Gesammelte Abhandlungen, E. Zermelo,
"Time and entropy," Am. Scientist 43, 550 (1955).
Ed. (Springer, Berlin, 1932).
8. For details on the relevant mathematical and phil- 15. P. W. Bridgman, Reflections of a Physicist (Phil-
osophic issues, see A. Griinbaum, "A consistent con- osophical Library, New York, 1950), pp. 162-167.
ception of the extended linear continuum as an ag- 16. See A. Griinbaum, "Relativity and the atomicity of
gregate of unextended elements.," Philosophy of becoming," Rev. Metaphysics 4, 143-186, esp. pp.
Science 19, 288 (1952). 176-184 (1950); "Messrs. Black and Taylor on
9. See E. W. Hobson, The Theory of Functions of a temporal paradoxes," Analysis 12, 146-148 (1952).
Real Variable, (Cambridge Univ. Press, Cambridge, 17. H. Weyl, Philosophy of Mathematics and Natural
ed. 2, 1921), vol. I, pp. 56-57. Science (Princeton Univ. Press, Princeton, N.J.,
10. See W. Hurewicz and H. Wallman, Dimension 1949), p. 43. For a discussion of the Democritean
Theory (Princeton Univ. Press, Princeton, N.J., conception of mathematical atomism and its re-
1941), p. 18, and, in reference 8, pp. 290-295. See ception in antiquity, see S. Luria, "Die Infinitesimal-
also P. S. Aleksandrov, "The present status of the theorie der antiken Atomisten," in Quellen u.
theory of dimension," Am. Math. Soc. Tranislations 2, Studien zur Gesch. d. Math. Astr. u. Phys., Abt. B:
1 1 (10rr,\ Studien, II (Springer, Berlin, 1933), pp. 106-185.

A great deal of the joy of life consists in doing perfectly, or at least to the best of
one's ability, everything which he attempts to do. There is a sense of satisfaction, a
pride in surveying such a work-a work which is rounded, full, exact, complete in all
its parts-which the superficial man, who leaves his work in a slovenly, slipshod, half-
finished condition, can never know. It is this conscientious completeness which turns
work into art. The smallest thing, well done, becomes artistic.-WILLIAM MATHEWS.

November 1955 23~9

This content downloaded from

90.166.56.191 on Tue, 11 Mar 2025 09:19:29 UTC
All use subject to https://about.jstor.org/terms