Most chromosome counts in the section are 2N = 30 though A. microspadix Schott has a count of 2N = 60 (Sheffer & Croat, 1983) and A. pulchellum Engl. has a report of 2N = 63 (Mookerjea, 1955) which probably represents the presence of aneuploidy in the section.

Few attempted interspecific hybridizations have been reported, however fruit set has been obtained in a pollination of A. testaceum  Croat & Baker x A. radi cans C. Koch (section Chamaerepium).

X. POLYNEURIUM Engl. (Fig. 18)
This group, which seems somewhat unnatural, is characterized chiefly by having relatively thin blades which usually have numerous, closely parallel primary lateral veins. The internodes may be short or elongate. The most well known examples include A. argyrostachyum Sodiro, A. corrugatum Sodiro, A. cuspidatum and A. panduriforme Schott. Blade shape in the section is highly variable, ranging from cordate to more or less oblong which may be acute to subcordate at the base. A few of the species in this group are well known ornamentals but most members of this relatively large group are poorly known.

The section is poorly known chromosomally with A. caperatum Croat & Baker having 2N = 30 and three counts for A. wallisii Masters of 2N = 30, 2N = 30 + 2B and 2N = 60. While there is no way to confirm the identities of the plants used in these counts it is assumed that at least the Sheffer and Kamemoto counts (1976) represented a member of section Cardiolonchium . There is growing at Kew Gardens a plant bearing the name A. wallisii  which looks much like A. rubrinervium (Link.) G. Don (a bona fide member of section Cardiolonchium ) and this plant in no way matches the type specimen which is a typical member of section Polyneurium.

Information on crossability in section Polyneurium is lacking, however fruit set was obtained in an attempted cross between A. caperatum x A. ravenii Croat & Baker (both sect. Polyneurium but with the latter having been reported erroneously in Croat, 1983 & 1983 as sect. Calomystrium).

XI. UROSPADIX Engl. (Fig. 19-21)
This is the largest group defined by Engler and as defined it is certainly the most unnatural assemblage in the genus. All of the glandular-punctate species (mostly described after Schott's time) are correctly placed in section Porphyrochitonium . Once this natural assemblage is removed, most of the remainder consists of a seemingly natural group of species concentrated in eastern and southeastern Brazil. It is this group which contains such examples as A. bellum Schott, A. binotii  Linden (Fig. 19), A. comtum Schott (Fig. 20), A. crassipes Engl., A. galeotii  (Hort.) C. Koch, A. lucidum  Kunth (Fig. 21), A. miguelianum  C. Koch & Augustin, A. olfersianum  Kunth and A. vittariifolium  Engl., that should be properly considered section Urospadix . As defined, section Urospadix exhibits relatively uniform characteristics. Generally species in this section have short stems with short internodes and generally epunctate leaf blades which are typically much longer than broad (being typically lanceolate). Only rarely are they cordate or subcordate at the base. Perhaps the most important characteristic is the typically close, numerous primary lateral veins which are scarcely more prominent than the interprimary veins.

11