THE MEGASPORANGIUM 71 



primary sporogenous cell usually functions as the megaspore mother 

 cell without undergoing any further divisions. 



The outline presented above is subject to many variations. In 

 some plants the archesporial cell is said to originate from the third 

 layer of cells in the nucellus, but this is probably a misinterpretation 

 caused by the difficulty in distinguishing the archesporial cell at 

 an earlier stage of development. Sometimes, as in the Onagraceae 

 (Khan, 1942), the archesporium may comprise a small group of 

 half a dozen cells or more (Fig. 5022). Of these usually the central 

 cell alone is functional, but frequently one or two of the other cells 

 also reach the megaspore mother cell stage. In the Malvaceae 

 (Stenar, 1925) the primary sporogenous cell divides to form a few 

 accessory cells in addition to the functional megaspore mother cell. 

 In some members of the Rubiaceae and Compositae there are sev- 

 eral sporogenous cells, all of which may go through the meiotic 

 divisions (Fig. 50B-C). In Scurrula (Rauch, 1936) and Dendroph- 

 thoe (Singh, 1950), which have a very massive archesporium, the 

 sporogenous cells undergo further division to give rise to a still 

 larger number of cells. These begin to elongate very actively and 

 become so closely interlocked that the whole tissue gives an appear- 

 ance suggestive of the hy menial layer of an ascomycete. 



In Hydrilla there are sometimes two or three archesporial cells 

 in a single row (Fig. 50A). In Ruppia (Murbeck, 1902), Butomus 

 (Holmgren, 1913), and Urginea (Capoor, 1937) the primary parie- 

 tal cell may also assume a sporogenous function so that two mega- 

 spore tetrads are formed in the same row (Fig. 52D). In Oncidium 

 praetextum (Afzelius, 1916) a cell of the nucellar epidermis may func- 

 tion as a megaspore mother cell (Fig. 50F), and in Solanum (Bha- 

 duri, 1932) and Limnanthes (Fagerlind, 1939) some of the integu- 

 mentary cells may behave similarly (Fig. 50D,G). 



In the Sympetalae a parietal cell is absent, the only important 

 exceptions being the Plumbaginales and some members of the fam- 

 ily Convolvulaceae. Since this is also the condition in some other 

 advanced families like the Umbelliferae and Orchidaceae, the pres- 

 ence of a massive parietal tissue is regarded as a primitive feature 

 and its absence as advanced. An objection to this view is that 

 parietal cells are often absent even in some admittedly primitive 

 families like the Ranunculaceae (Hafliger, 1943). 



In the Casuarinaceae (Fig. 51), and some other families (see 



