Like many other genera of hardy trees and shrubs, the catalpas are found in both the Old and New Worlds. Although first made known to English cultivators from N. America in the form of C. bignonioides (which was introduced in 1726), the genus has been found in later times to be more abundantly represented in China, where four to six species occur. In the open ground the catalpas form low, wide-spreading, bushy-headed deciduous trees, a habit largely due, no doubt, to the shoots never forming a terminal bud. The young wood is stout, and very pithy. Leaves either opposite or in threes, large, long-stalked. Flowers produced in panicles, corymbs, or racemes at the end of the shoots of the year. Corolla bell-shaped at the base, with five spreading, frilled lobes; calyx two-lipped; stamens five, only two of which as a rule are fertile. Seed-vessel a very slender, cylindrical capsule 1 to 2 ft long, and 1⁄4 to 1⁄2 in. diameter. Seeds numerous, flat, with a fringe of long, white hairs at each end.
Three Asiatic species have been introduced but none has proved superior to the American catalpas C. bignonioides and speciosa. These are undoubtedly among the most beautiful of all flowering trees and an isolated tree on a lawn is seen to exceptional advantage. The hybrid C. × erubescens 'Purpurea' also makes a very striking specimen. At the same time all the species are worth cultivation.
Catalpas like generous treatment at the root; a deep, moist loam is best, and an open, sunny, but not a bleak spot. Owing to the branches never forming a terminal bud and the annual bi- or tri-furcation this induces, it is advisable when the trees are young to train up a leader high enough to produce a trunk of the desired height, say 10 ft, when the tree may be left to assume the spreading habit natural to it. C. bignonioides thrives well in London, and for many years there have been fine specimens in Parliament Yard, Westminster. For propagation I would prefer seeds to any other means, believing that trees so raised are the longest-lived. But when these are unobtainable, and for distinct forms or coloured-leaved varieties, cuttings may be used. These should be made of the young leafy shoots as soon as they are moderately firm, and struck in mild bottom-heat.
The most recent treatment of the genus is by the Czechoslovak botanist Jiri Paclt in Candollea, 1952, pp. 241-85.