Let us define by induction a countable family (In,p) of mutually disjoint open intervals as follows: The integer n takes all values ⩾0, and for each value of n, p takes the values 1,2,3,…,2n. All the intervals In,p are contained in
A=[0,1], and we take
I0,1=]1/3,2/3[ (the "middle third" of
]0,1[). Suppose now that the
2m+1−1 intervals
In,p have been defined for
0≤n≤m in such a way that, if
Jm is their union, the set
A∩∁Jm is the union of
2m+1 mutually disjoint closed intervals
Km,p (
1≤p≤2m+1) each of length
3m+11. If
Km,p=[a,b] we then take
Im+1,p to be the open interval
[a+3b−a,3b−a] (the "middle third" of the interval
]a,b[), and it is immediately verified that the induction can continue in this way (Fig. 3).
!Diagram showing intervals I_{2,1}, I_{1,1}, I_{2,2}, I_{0,1}, I_{2,3}, I_{1,2}, I_{2,4} and corresponding intervals K_{2,1}, K_{2,2}, K_{2,3}, K_{2,4}, K_{1,1}, K_{1,2}, K_{1,3}, K_{1,4}, K_{0,1}, K_{0,2}
Figure 3.
If K′ is the complement of the union of the In,p, the closed set K=A∩K′ is called Cantor's triadic set. Clearly K is compact (no. 2, Theorem 2); also K is totally disconnected. For if K contained an interval I of length >0, then I would be contained in some interval Km,p; hence its length would be ≤1/3m+1 for each m, which is absurd.
Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.
Nothing.