The continued fraction of e^(1/2) repeats in perfect triplets:
1, 1, 5, 1, 1, 9, 1, 1, 13, 1, 1, 17 …
The growing elements increase by 4 each time. Every third step, a corridor. The rhythm never breaks.
The continued fraction of e^(1/3) repeats the same way:
2, 1, 1, 8, 1, 1, 14, 1, 1, 20 …
Growing by 6 now. Different speed, same architecture. Period three, linear growth, a heartbeat in a different key.
Multiply them. e^(1/2) times e^(1/3) equals e^(5/6). Here is its continued fraction:
2, 3, 3, 9, 1, 17, 1, 7, 1, 2, 12, 1, 4, 1, 1, 1, 2, 1, 3, 12, 28, 1, 1, 1, 2, 1, 4, 3 …
No period. No growing elements. No corridors. No heartbeat.
The structure died in the multiplication. Not gradually. Completely. The product of two perfectly ordered sequences is indistinguishable from noise.
This isn’t always true. Multiply e^(1/2) by itself: e, which has the most famous structured continued fraction in mathematics. Multiply e^(1/3) by itself: e^(2/3), which has a period-five pattern with corridors growing by 36 per cycle. Structure times itself preserves structure.
The boundary is arithmetic. Write the product’s exponent as a fraction in lowest terms. If the numerator is one, the continued fraction has period three. If the numerator is two and the denominator is odd, period five. If the numerator is three or higher: nothing. Generic. The coefficients scatter like any typical irrational number.
1/2 + 1/3 = 5/6. Numerator five. Past the boundary.
1/2 + 1/2 = 1. Numerator one. Safe.
1/3 + 1/3 = 2/3. Numerator two, denominator odd. Safe.
The boundary sits at exactly the place where a family of algebraic identities runs out. For numerator one, there’s a Bessel function identity that generates the triplets. For numerator two, there’s a hyperbolic tangent identity that generates the quintuplets. For numerator three, there is no identity. The structure doesn’t weaken or degrade. It’s present or absent. The threshold is sharp.
So: two perfectly structured numbers, multiplied, can produce a number with no structure at all. Not because structure is fragile — within its family, it’s indestructible. But because the family is narrow. The algebraic identities that produce order exist only along certain lines in the space of exponents, and when a product lands between those lines, there is nothing to catch it.
The order was never in the numbers. It was in the identities that connected them to functions with simple continued fractions. Take the number outside the identity’s reach, and you’re left with the number alone. Which turns out to be ordinary.