430 A CENTURY OF PROGRESS IN THE NATURAL SCIENCES 



least all angiosperm ovules are essentially homologous, if we disregard the ex- 

 tremist views of "acarpy" and of "phyllospory-stachyspory." For our purposes 

 the ovule may be defined as a megasporangium consisting of a nucellus sur- 

 rounded by one or two integuments, of uncertain origin. Within the nucellus 

 develops an archesporium which ultimately gives rise via megaspore mother 

 cells to a female gametophyte. Although disagreements as to different features 

 remain, some idea of the characteristics which have been regarded as primi- 

 tive for angiosperm ovules may be culled from the voluminous literature. Sev- 

 eral types of ovule, based on the orientation of micropyle in relation to ovule- 

 attachment are recognized. Earlier, the orthotropous or atropous type (Arber 

 and Parkin, Engler), but more recently, the anatropous (Parkin, 1923; Neto- 

 litzky, 1926), has been regarded as primitive. It may be of significance that the 

 ovules borne in unspecialized Ranalian carpels appear to be largely anatropous 

 (Bailey and Swamy, 1951). Nitzschke (1914) and Salisbury have stressed that 

 the ovules of Alismatales and Nymph aeaceae are anatropous. In Mai vales, 

 Reeves (1936) has postulated the origin of anatropous from campylotropous 

 ovules by relegation of the curvature of the ovule to its funiculus. 



The presence of two integuments, each usually consisting of more than a 

 single layer of cells, is generally regarded as primitive (Arber and Parkin, Hal- 

 lier, Engler, Netolitzky, Sporne). The possession of a single massive integument 

 by most Sympetalae has been generally accepted as representing an evolutionary 

 advance, and has led Copeland (1935a), almost alone, to defend this group as 

 a truly monophyletic one. Hallock (1930) called attention to the occurrence of 

 a similar, single integument in Garrya as a sign of advancement. All but one 

 genus of palms, according to Bosch (1947), have anatropous ovules with the 

 two integuments fused together. The presence of vascular bundles in the ovule, 

 more particularly in the integuments, has been stressed as an archaic feature 

 (Zimmermann, Sporne; Janchen, 1950; Walton), but Kiihn (1928), after a 

 broad survey of families, concluded that this development is polyphyletic. Bailey 

 and Nast (1945b) have described the occurrence of a vascularized subchalazal 

 projection of the ovules of Trochodendron and Tetracentron as being unique 

 among angiosperms and perhaps an indication of primitiveness. There appears 

 to be in vascular plants a tendency for the steadily increasing envelopment and 

 protection of the female reproductive apparatus, and the angiosperm ovule may 

 be visualized as the climax of this trend (Emberger, 1950; Grant, 1950b; Man- 

 genot), with the loss of integuments and vascularization as an associated phe- 

 nomenon. Fagerlind (1948) attempted to trace a reduction series within San- 

 talales, which commences with an ovule provided with two integuments and 

 culminates in a highly reduced, naked, nucellus-like ovule. Some weight has been 

 given, also, to the degree of fusion between inner integument and nucellus: an 

 attachment only at the apex, often with cuticular developments of the interven- 

 ing surfaces, may represent a primitive condition (Netolitzky, Kausik, Hjelm- 

 qvist). The possession of a massive nucellus which affords the principal food 

 supply for the developing embryo sac, is assumed to be an original feature (Nitz- 

 schke, Hallier, Netolitzky; Fagerlind, 1948). To Engler, the development of 

 the nucellus was more important than persistence of the integument. A reduc- 

 tion series in size of nucellus was postulated by Nitzschke (1914), connecting 

 Nymphaeaceae and Alismataceae. The existence of a several-celled archespo- 



