You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: index.html
+56-84Lines changed: 56 additions & 84 deletions
Original file line number
Diff line number
Diff line change
@@ -5062,142 +5062,112 @@ <h4>Archimedes and the Illusion of Limits</h4>
5062
5062
</mrow>
5063
5063
</math>
5064
5064
<br><br><br>
5065
+
5066
+
<pitemprop="disambiguatingDescription">If the true circumference is 6.4r, then:</p>
5065
5067
<pitemprop="disambiguatingDescription">But the entire method relies on the unproven assumption that every polygon produced by repeated angle bisection of a circumscribed polygon remains circumscribed.
5066
5068
<br>
5067
-
This is never demonstrated. It is simply taken for granted.</p>
5068
-
<br><br><br>
5069
+
This is never demonstrated. It is simply taken for granted.
5070
+
<br>
5071
+
And the assumption is false once the polygon has too many sides.
<li>A tangent segment is always longer than the arc it touches.</li>
5073
-
<li>A polygon with its perimeter close enough to the circumference cannot remain outside the circle. It must cross the arc.</li>
5074
-
</ul>
5075
+
The math world loves the idea that a regular polygon with a side count approaching infinity “collapses into a circle.” It’s a charming picture — but it is highly theoretical and ultimately unrealistic.
5076
+
<br><br>
5077
+
When the perimeter of a polygon equals the circumference, its sides are too short to remain tangent to the circle. They cut through the arc. This behavior is easy to see in polygons with a low side count.</p>
<imgclass="center-fit" src="isoperimetry.png" alt="The sides of a hexagon cross the arc of an isoperimetric circle">
5078
5081
</figure>
5079
5082
<br>
5080
-
<pitemprop="disambiguatingDescription">This is not optional; it is a geometric necessity. The only curve with its perimeter exactly equal to the circumference that never intersects its interior disk is the circle itself.
5081
-
<br><br>
5082
-
The bisection procedure guarantees that each new perimeter is smaller than the previous one.
5083
-
<br>
5084
-
But whether a polygon produced by angle bisection remains tangent depends on whether its perimeter is still sufficiently greater than the circumference.
5085
-
<br><br>
5086
-
In conventional geometry, a polygon is called circumscribed if each side touches the circle at exactly one point.
5083
+
<pitemprop="disambiguatingDescription">In conventional geometry, a polygon is called circumscribed if each side touches the circle at exactly one point.</p>
<li>A hexagon can be circumscribed about a circle.</li>
5090
5087
<li>An octagon can still be circumscribed.</li>
5091
-
<li>But as the number of sides increases, each side becomes shorter and shorter, and eventually the straight sides can no longer maintain true tangency. They begin to cut through the circle instead of just touching it.</li>
5088
+
<li>But as the number of sides increases, each side becomes shorter and shorter, and eventually the straight sides can no longer maintain true tangency. They begin to cut through the circle instead of just touching it.
5089
+
</li>
5092
5090
</ul>
5091
+
<br>
5092
+
<pitemprop="disambiguatingDescription">Whether a polygon can be circumscribed about a circle depends on whether its perimeter is sufficiently greater than the circumference.
5093
5093
<br><br>
5094
-
<pitemprop="disambiguatingDescription">Think of a honeycomb cell.
5094
+
But the bisection procedure guarantees that each new perimeter created by repeatedly bisecting a circumscribed hexagon is smaller than the previous one.
5095
5095
<br><br>
5096
-
The bees build a perfect hexagon, fill it, and later dig a circular hole inside it.
5096
+
To know if the polygons are still circumscribed at every finite step, you must already know if the perimeter is large enough that no crossing occurs.
5097
5097
<br>
5098
-
If we treat the filled space as a perfect hexagon and the hole as a perfect inscribed circle, we get a very clear picture of what “circumscribed” and “inscribed” really mean in practice.</p>
<li>Bees could make an octagon around a circle.</li>
5103
-
<li>But at some point — around 12 sides — the walls would collapse inward because the straight edges cannot maintain tangency.</li>
5104
-
</ul>
5105
-
<br>
5106
-
<pitemprop="disambiguatingDescription">This is why Archimedes’ method fails:
5107
-
<br>
5108
-
it assumes that each side of every polygon with a finite side count produced by repeatedly bisecting the circumscribed hexagon must touch the circle at exactly one point.
5109
-
<br>
5110
-
But this is never proven — it is simply assumed.
5111
-
<br>
5112
-
And the assumption is false once the polygon has too many sides.
5113
-
<br>
5114
-
A straight line can always be drawn, but that does not guarantee single-point contact.
5115
-
<br>
5116
-
When the perimeter of the polygon approaches the circumference, the sides will either touch along a small arc or cut into the circle.
5117
-
<br>
5118
-
The failure is not the existence of the line — the failure is the loss of single-point tangency.
5098
+
Yet finding the circumference is the purpose of the method. The method therefore relies on the very assumption it seeks to prove. That is circular reasoning.
5119
5099
<br><br>
5120
-
The standard argument claims the polygons remain outside forever at a finite side count and only collapse into the circle at infinity. But this is circular reasoning. To know if the polygons are still circumscribed at every finite step, you must already know if the perimeter is large enough that no crossing occurs.
5121
-
<br>
5122
-
Yet finding the circumference is the purpose of the method. The method therefore relies on the very assumption it seeks to prove.
5123
-
<br>
5124
-
You cannot use the method to justify the assumption that makes the method valid.</p>
5100
+
You cannot use the method to justify the assumption that makes the method valid.
5101
+
<br><br>
5102
+
The classical argument cannot prove that the polygon remains tangent, and without that proof, it cannot be used to refute the circumference C = 6.4r.
5125
5103
<br><br>
5126
-
<pitemprop="disambiguatingDescription"><b>The breakdown is visible at small finite n when C = 6.4r.</b>
5104
+
The flaw in the classical method becomes even clearer when we try to implement it practically.
5127
5105
<br><br>
5128
-
The area–square construction gives the true value.
5106
+
The breakdown is visible already at a small finite side count when attempting to draw a circumscribed 24-gon or 48-gon via exact angle bisection (central angle 15° → 7.5°).
5129
5107
<br>
5130
-
The quadrant-to-square rearrangement with uncovered area = total overlap area yields a square of area exactly 3.2r².
5108
+
Tthe figure collapses:
5131
5109
<br>
5132
-
Since the four quadrants are the original pieces of the circle, the circle area is exactly 3.2r².
5110
+
The supposed tangent lines merge, overlap, or cross the arc, behaving like an isoperimetric polygon rather than a circumscribed one.
5111
+
<br>
5112
+
This is not a precision or rendering error; it is the geometry refusing to produce a valid set of external tangents.</p>
5133
5113
<br><br>
5134
-
Differentiating with respect to radius gives circumference exactly 6.4r.
5114
+
<pitemprop="disambiguatingDescription">The area–square construction gives the true ratio.
5115
+
<br>
5116
+
Differentiating with respect to the radius gives a circumference exactly 6.4r.
5135
5117
<br>
5136
5118
This is not an approximation; it follows directly from finite geometric construction.
5137
5119
<br>
5138
5120
This finite construction avoids the infinite-regress trap entirely.
5139
-
<br>
5140
-
The flaw in the classical method becomes even clearer when we try to implement it practically.
5141
5121
<br><br>
5142
-
When attempting to draw a circumscribed 24-gon or 48-gon via exact angle bisection (central angle 15° → 7.5°), the tangent lines merge, overlap, or cross the arc — even in high-precision vector software. The individual sides become indistinguishable or intersect the circle before reaching distinct tangent points. This is not a precision or rendering error; it is the geometry refusing to produce a valid set of external tangents.</p>
<imgclass="center-fit" src="polygonApproximation.png" alt="When attempting to draw a circumscribed 24-gon via exact angle bisection (central angle 15°), the tangent lines merge, overlap, or cross the arc. The individual sides become indistinguishable or intersect the circle before reaching distinct tangent points.">
5146
5128
</figure>
5147
5129
<br>
5148
-
<pitemprop="disambiguatingDescription">A circumscribed n‑gon has perimeter:
<li>The 12‑gon is a borderline case. Its perimeter is still slightly greater than 6.4r, so it may or may not maintain true tangency.</li>
5163
5141
<li>But the 24‑gon — and any polygon with more sides — cannot be circumscribed at all, because a circumscribed polygon must always satisfy P(n) > C. Once P(n) drops below 6.4r, the sides become too short to touch the circle at a single point. They inevitably cut through the arc instead of remaining outside it.</li>
5164
5142
</ul>
5165
5143
<br>
5166
-
<pitemprop="disambiguatingDescription">The construction no longer produces a proper set of distinct tangent sides — it fails in a literal, physical sense. The required tangent lines from adjacent vertices converge so sharply that they overlap or intersect the arc before reaching distinct tangent points.
5144
+
<pitemprop="disambiguatingDescription">Think of a honeycomb cell.
5145
+
<br><br>
5146
+
The bees build a perfect hexagon, fill it, and later dig a circular hole inside it.
5147
+
<br>
5148
+
If we treat the filled space as a perfect hexagon and the hole as a perfect inscribed circle, we get a very clear picture of what “circumscribed” and “inscribed” really mean in practice.</p>
5167
5149
<br>
5168
-
The figure collapses: instead of clear external tangents, the tangent lines for a 24‑gon or 48‑gon merge, overlap, or cut through the circle, behaving like an isoperimetric polygon rather than a circumscribed one.
The assumption “we can always bisect again and obtain another circumscribed polygon” fails constructively — at a small, finite number of sides, not only in some unreachable infinite limit.
5171
-
<br><br><br>
5172
-
<b>The Consequence</b>
5159
+
<br>
5160
+
Once the perimeter of the constructed polygon falls below the true circumference, the tangent‑doubling formulas no longer describe a real circumscribed polygon. They describe a figure that has already slipped inside the circle.
5173
5161
<br><br>
5174
-
Once the perimeter of the constructed polygon falls below the true circumference, the tangent‑doubling formulas no longer describe a real circumscribed polygon. They describe a figure that has already slipped inside the circle. The apparent convergence toward ~6.28r is therefore not a discovery of the pi, but an artifact of continuing a construction that has already become geometrically impossible.
5162
+
So the theoretical “infinite‑sided polygon” does not resemble a circle at all.
5163
+
<br>
5164
+
It is simply a polygon whose sides have become so short that they no longer maintain tangency. It behaves more like a polygon pushed inward by isoperimetry than like a circle approached by limits.
5175
5165
<br><br>
5176
-
The classical argument cannot prove that the polygon remains tangent, and without that proof, it cannot be used to refute the circumference C = 6.4r.
5166
+
Hence, the convergence toward ~6.28r is not a discovery of the pi, but an artifact of forcing a construction that has already become geometrically impossible.
5177
5167
<br><br>
5178
5168
These structural issues in the polygon‑limit method set the stage for a second misconception: the symbolic fusion of an approximation with the geometric ratio it was meant to represent.</p>
5179
5169
</section>
5180
5170
<br><br><br>
5181
-
<sectionid="symbol">
5182
-
<h4>The Symbol π: A Linguistic Shortcut</h4>
5183
-
<br><br>
5184
-
<pitemprop="disambiguatingDescription">Since the numeric result of Archimedes’ approximation was an infinite fraction — 3.14…— whose digits cannot all be written, he needed a symbolic notation for it in his formulas.
5185
-
<br><br>
5186
-
Technically, the circumference is a perimeter. So the perimeter‑to‑diameter ratio — P / d — became pi over delta in Greek. With the diameter chosen as the reference — d = 1 —, this simplifies to pi / 1 = pi.
5187
-
<br><br>
5188
-
<strong>But this is not necessarily the ratio itself — it is the notation of that ratio. That distinction matters.</strong>
5189
-
<br><br>
5190
-
The circumference‑to‑diameter ratio is a universal geometric proportion.
5191
-
<br><br>
5192
-
The numeric value commonly associated with it — 3.14…— is the result of Archimedes’ polygon‑based approximation method.
5193
-
<br><br>
5194
-
<strong>This is how a numerical output of a failed computational estimate gradually hardened into a symbol, and the symbol into a “geometric constant.” </strong>
5195
-
<br><br>
5196
-
Over time, the numerical result of Archimedes’ polygon‑limit procedure was reinterpreted as a fundamental property of the circle itself.
5197
-
<br><br>
5198
-
It was not until the 18th century that the symbol, popularized by the mathematicians of the time, gained widespread acceptance, meanwhile it took on a life of its own.</p>
5199
-
</section>
5200
-
<br><br><br>
5201
5171
<sectionid="calculus">
5202
5172
<h4>∫ Calculus: Summary, Not Source</h4>
5203
5173
<br><br>
@@ -5797,7 +5767,9 @@ <h4 style="margin:12px" itemprop="disambiguatingDescription">The widely used for
5797
5767
<br><br>
5798
5768
This is an identity; not tautology.
5799
5769
<br><br>
5800
-
Circle area = square area</p>
5770
+
The quadrant-to-square rearrangement with uncovered area = total overlap area yields a square of area exactly 3.2r².
5771
+
<br>
5772
+
Since the four quadrants are the original pieces of the circle, the circle area is exactly 3.2r².</p>
0 commit comments