116 MORPHOLOGY OF THE ANGIOSPERMS 



side of the sporophyll comes also from somewhat less primitive families, 

 the Magnoliaceae and Nymphaeaceae, which have stamens transitional 

 in many characters between laminar and anther-filament types. In these 

 and other families, the relation of the margin to sporangium position is 

 obvious. 



An important aspect of sporangium position on the sporophyll seems 

 to have been overlooked. The laminar stamens of some families — De- 

 generiaceae, Himantandraceae, Calycanthaceae, Annonaceae, Ceratophyl- 

 laceae, Lactoridaceae, Lardizabalaceae — have the sporangia on the 

 abaxial side; sporangia are on the adaxial side in other families — Mag- 

 noliaceae, Nymphaeaceae, Austrobaileyaceae (Fig. 50). Rarely, on semi- 

 laminar stamens, the four sporangia have a marginal or submarginal 

 position — Drimijs. The position of sporangia in the most primitive sta- 

 mens—in some taxa, adaxial (ventral); in others, abaxial (dorsal)— is 

 morphologically puzzling. In vascular plants, sporangium position has 

 been generally accepted as an important character of major taxa. The 

 angiosperms appear to have, primitively, sporangia in both positions. 



In outline, the evolutionary history of the stamen seems to have been 

 much as follows. The primitive stamen was laminar, with two pairs of 

 sporangia borne on either the adaxial or the abaxial surface. From tliis 

 simple stamen has been developed the slender, complex organ with 

 marginal pairs of sporangia. Dorsiventral form has been largely lost; 

 the specialized anther of the higher families is more or less four-angled 

 and the filament terete. The marginal position of the pairs of sporangia 

 — one member of each pair apparently adaxial, the other abaxial — sug- 

 gests that the sporangia belong two on one side and two on the other, 

 but, in the primitive stamen, all four are on one side (Fig. 50). The 

 marginal position of the pairs of sporangia is clearly derived, but the 

 story of change in location is involved. 



Sporangium position in the anther was much discussed between 1820 

 and 1850, but with little agreement as to its interpretation. The view 

 at that time was that all sporangia belong on the adaxial side, but 

 there was "great doubt whether the anther chambers have arisen where 

 they appear"; "whether two sporangia do not belong on one side and two 

 on the other"; "whether the manifold structure of anthers permits . . . 

 the assumption of one type for all stamens." These questions might be 

 raised today, but an acquaintance with the primitive stamen type, which 

 was unknown to earlier morphologists, now aids greatly in the interpre- 

 tation of this obscure situation. 



In one interpretation of these early days, the stamen was considered 

 inroUed, like the carpel, with its inner, enclosed surface sporogenous, 

 and with the pollen shed through breaks in the sporophyll wall. (That 

 all the sporangia are on one side of the sporophyll is implied.) This 



