Die Zahlen 33, 65, 138 geben ganz günstige Annäherungswerthe an die Verhältnisszahlen 33, 66, 132.
The numbers 33, 65, 138 give quite favourable approximate values to the numerical proportions of 33, 66, 132.

It seems a strange coincidence that Mendel ends up with multiples of 33. This result suggests itself quite strikingly already in the “ordered” table presenting the empirical results on the same page above this sentence (see p. 20).

numerical proportions = Zahlenverhältnisse See p. 3, s. 8.