The Koch snowflake, and how its sides and perimeter grow at each step

What to assume they know

• Self-similarity: a shape that contains copies of itself — self-similarity and the step-sequence idea
• The Sierpinski carpet and gasket: what repeated removal leaves behind — how a counting recursion is read off a construction, and how a recursion becomes a closed form
• Dividing a segment into three equal parts, and the idea of one-third of a length
• Equilateral triangles and their 60° angles
• Perimeter as the total boundary length of a closed figure
• Multiplying and comparing fractions, enough to see that 4/3 > 1
• Powers with whole-number exponents

What they should be able to do

• State the snowflake's construction rule in the two steps the chapter gives, and restate it as "each side becomes a bump"
• Count the segments a single side is replaced by, and give the reason there are four rather than three
• Derive the number of sides at step n as 3 × 4ⁿ, and justify the factor of 4 by an argument about one side
• Derive the length of one side at step n as (1/3)ⁿ of the starting side
• Combine the two to get the perimeter at step n, taking the starting side as 1 unit
• Explain why multiplying by 4/3 repeatedly makes the perimeter exceed any figure you care to name
• Explain why the figure nevertheless stays inside a bounded region, and identify a region that contains every step
• Contrast the snowflake with the two Sierpinski fractals: adding versus removing, and a boundary that lengthens versus an area that shrinks

Where it usually goes wrong

• "Each side is divided into three, so there are three new segments." Four. The middle third is deleted and two sides of the raised triangle take its place, so 3 − 1 + 2 = 4. Counting this wrongly is the single most common error in this construction, and it changes every subsequent number.
• "The perimeter grows by adding one-third each time, so it grows slowly." It grows by one-third of its current value, which is multiplication, not addition. Repeated multiplication by 4/3 outruns any fixed target: 30 steps in, a snowflake starting from a triangle of side 1 cm has a boundary of roughly 170 m — a boundary the length of two football pitches, folded into a figure still small enough to cover with a hand.
• "An unbounded perimeter means the shape gets bigger and bigger." It does not. Every step fits inside the same small circle. The boundary gets longer by becoming more crinkled, not by spreading out.
• "If the perimeter is unbounded, the area must be too." The two are governed by different powers of the shrink factor. Say it with the 4 × 1/3 against 4 × 1/9 comparison.
• "Step 1 is a hexagon." It is a six-pointed star — twelve sides, not six. A student who reads Step 1 as a hexagon gets S₁ = 6 and everything afterwards is wrong. Count the sides.
• "The bump points inward." Outward, always, and the chapter's Step 1 star shows it. Pointing inward gives a different (and also interesting) fractal, but not this one, and it is worth naming as a wrong turn rather than leaving it as a silent assumption.
• "This is the same kind of construction as the Sierpinski ones." Same machine, opposite move. Sierpinski removes material and the area falls; Koch adds material and the boundary lengthens. Putting the two side by side is what makes either of them mean anything.

Questions to check understanding

• Count the number of sides at a stated step and justify the count
• Find the length of one side at a stated step, given the starting sidelength
• Compute the perimeter at a stated step for a starting side other than 1 unit
• Decide, with a reason, whether repeated multiplication by a given factor makes a quantity grow without bound
• Given a variant rule — for example, dividing each side into five parts — rework the side count and perimeter
• Explain how a figure can have an ever-lengthening boundary while staying inside a fixed region
• Draw Step 2 of the sequence accurately, which is the chapter's own item 1

Examples worth working on the board

Values marked not in the book are worked out here on the chapter's stated inputs; the chapter prints no answers.

• Attribution, as printed (Part II p.73). The fractal carries the name of Von Koch, a mathematician from Sweden, whose first description of it the chapter dates to 1904. The chapter also states that this fractal was already met in Class 6 Ganita Prakash — the only cross-grade pointer in §4.1.
• The rule, in the two numbered moves the chapter splits it into (Part II p.73). Begin with an equilateral triangle. Move (i): cut every side into three equal pieces. Move (ii): on the middle piece build an equilateral triangle standing outward, then rub out that middle piece. The chapter's summary of the pair is that a side has in effect been swapped for a bump-shaped structure, and a small drawing of that bump is set inline in the sentence. Then do the same to every side of the figure you now have.
• The printed step figures (Part II p.73, artwork). Step 0 is a plain equilateral triangle. Step 1 is a six-pointed star. Step 2 is the star with a bump on each of its edges. An ellipsis follows. Below them sits a near-final picture captioned Koch Snowflake. All four are drawn in green on the printed page.
• The exercise set (Part II p.73, Figure it Out, three items). Item 1: draw the first steps of the sequence, at least to Step 2. Item 2: find the number of sides at step n. Item 3: find the perimeter at step n, and here the item fixes the scale — the triangle you start from has sides of 1 unit. That stipulation is printed and is the only quantitative input the section supplies.
• Not in the book, the side count. One side is replaced by four segments — the two outer thirds, plus the two sides of the raised triangle. The middle third is gone, so it contributes nothing. Every side is treated the same way and independently, so Sₙ₊₁ = 4Sₙ. With S₀ = 3 this gives S₁ = 12, S₂ = 48, S₃ = 192, and in general Sₙ = 3 × 4ⁿ.
• Not in the book, the side length. Each new segment is one of the thirds, so Lₙ₊₁ = Lₙ / 3. With L₀ = 1 this gives L₁ = 1/3, L₂ = 1/9, L₃ = 1/27, and in general Lₙ = (1/3)ⁿ.
• Not in the book, the perimeter. Pₙ = Sₙ × Lₙ = 3 × 4ⁿ × (1/3)ⁿ = 3 × (4/3)ⁿ. So P₀ = 3, P₁ = 4, P₂ = 16/3 ≈ 5.33, P₃ = 64/9 ≈ 7.11, P₄ = 256/27 ≈ 9.48. Ten steps in, P₁₀ = 3 × (4/3)¹⁰ ≈ 53.3; twenty steps in, about 946. An independent check on that formula: treat the step as an edit to the boundary rather than a recount. Each side loses its middle third and gains two segments of the same length, a net gain of one-third of a side per side — so the whole perimeter gains one-third of itself, which is multiplication by 4/3. On Step 0's three unit sides that gives 3 + 3 × (1/3) = 4, matching P₁.
• Not in the book, the containment argument for section 9. Every bump is raised outward, and the triangle raised on a segment of length L has height (√3/2)L < L, so no point of the new figure sits further from the old boundary than one current side length. Adding those bounds over all steps gives a total outward creep of less than 1/3 + 1/9 + 1/27 + … = 1/2 of the starting side. So a circle drawn around the starting triangle with a margin of half a unit contains every step, however many you run. The chapter makes no such argument; it is the reason the explanation can assert boundedness rather than assume it.
• Not in the book, the scaling reason. Shrink a shape by a factor k and its lengths shrink by k while its areas shrink by k². The snowflake multiplies the piece count by 4 while shrinking each piece by 1/3: for length that gives 4 × 1/3, which exceeds 1, but for area it would give 4 × 1/9, which is less than 1. The same rule grows the boundary and does not blow up the interior, and this single comparison is the cleanest statement of the thesis.
• **Values an explanation should not claim.** The area enclosed by the finished snowflake, the length "infinity", and the word dimension. None of the three appears in this chapter — checked against every page image — and none is Class 8 work.

Figures to have open

• The one-segment construction, shown in three beats: thirds marked, triangle raised, middle removed. The chapter prints only the tiny inline bump glyph (Part II p.73), so this must be drawn.
• The four step panels from Part II p.73, redrawn as schematics: triangle, star, bumped star, and a many-step picture. Step 1 must be countable — twelve sides.
• A single bump with its four segments numbered 1 to 4. Standard schematic; the chapter numbers nothing.
• A containment figure: the starting triangle, a circle around it with a marked half-unit margin, and three successive steps drawn inside. Standard schematic; the chapter draws nothing like it and section 9 depends on it.
• A perimeter table or bar chart for steps 0 to 5. Standard schematic.
• A side-by-side of a Sierpinski step and a Koch step for section 10, both redrawn.

Where this sits in the book

• NCERT Ganita Prakash Class 8, Part II, printed Chapter 4, §4.1 "Fractals", printed subheading "Koch Snowflake", Part II p.73 — the whole topic sits on that single printed page.
• The Figure it Out set is Part II p.73, items 1–3; item 3 supplies the 1-unit starting sidelength.
• Part II p.102, SUMMARY, bullet 2 lists the Koch Snowflake among the chapter's mathematical fractals and states that such fractals arise from repeatedly applied geometric operations.
• Cross-grade pointer stated on Part II p.73: this fractal appears in Class 6 Ganita Prakash.