**The most counterintuitive finding is that Iris-setosa has the widest sepals but by far the smallest petals.** Size does not scale consistently across the flower's parts. The per-species averages come from the `Iris_by_Species` gold table (50 specimens per species): - **Sepal width:** setosa is widest at about 3.42 cm. Virginica averages 2.97 cm and versicolor 2.77 cm, so setosa is about 0.65 cm wider than versicolor. - **Petal length:** setosa is smallest at about 1.46 cm, against 4.26 cm for versicolor and 5.55 cm for virginica. Its petals are roughly a third the length of versicolor's. - **Petal width:** setosa averages about 0.24 cm, against 1.33 cm for versicolor and 2.03 cm for virginica. - **Sepal length:** setosa is also the shortest at about 5.01 cm, against 5.94 cm for versicolor and 6.59 cm for virginica. Setosa is the smallest species on three of the four measurements, yet it is the widest on the fourth. If you assumed bigger flowers are bigger everywhere, sepal width would mislead you. Versicolor is narrowest at 2.77 cm but sits in the middle on the other three measurements, so sepal width does not follow the size ordering at all. The same pattern also separates setosa cleanly. Its largest petal length (1.9 cm) is below the smallest for any versicolor (3.0 cm), and its largest petal width (0.6 cm) is below versicolor's smallest (1.0 cm). Petal size alone therefore isolates setosa perfectly. Versicolor and virginica are not as clean: their petal ranges touch (versicolor's petal length max is 5.1 cm against virginica's min of 4.5 cm), and their sepal length ranges overlap heavily (both start at 4.9 cm). Sepal width is a poor guide to species, while petal size is a strong one. These figures are species-level minimums, maximums and averages. I did not test overlap at the level of individual specimens.