EMBRYOLOGY IN RELATION TO TAXONOMY 361 



parietal cells in the ovule; (7) embryo sac monosporic and eight- 

 nucleate, broader at the micropylar end and narrower at the chala- 

 zal; (8) a fluted hollow style, which connects the lumen of the ovary 

 with the outside and along which the pollen tubes make their way 

 into the ovary; (9) endosperm cellular; the first two divisions trans- 

 verse, giving rise to a row of four superposed cells (Fig. 198A-C); 

 (10) formation of endosperm haustoria at both ends of the embryo 

 sac, micropylar as well as chalazal (Fig. 198D, E); (11) a single- 

 layered seed coat formed from the outermost layer of the integu- 

 ment, the remaining layers becoming absorbed during the growth 

 of the endosperm and embryo; (12) seeds albuminous with fleshy 

 endosperm and straight embryo. 



All these are perfectly standard stages in Ericean embryology, 

 and their combination is quite unknown in any other order. The 

 Empetraceae show a close correspondence in all respects, while the 

 Sapindales and Celastrales differ in so many ways that there is no 

 doubt as to the correctness of Samuelsson's view. 10 



Lennoaceae. On the basis of his morphological studies on the 

 Lennoaceae, Solms-Laubach (1870) felt convinced that it belonged 

 to the Ericales, and Hutchinson (1926) accepted this disposition of 

 the group. But there were certain points in Solms-Laubach 's own 

 descriptions which seemed to militate against this view, and conse- 

 quently Engler and Gilg (1924) removed the Lennoaceae to the 

 order Tubiflorae and placed it in the neighborhood of the Boragi- 

 naceae. This received support from Siissenguth's (1927) anatomi- 

 cal and morphological study of Lennoa, and more recently Copeland 

 (1935) has also expressed his agreement with it on the basis of his 

 work on Pholisma. As remarked by Copeland, at the very outset 

 the equality in number of their stamens and corolla lobes (con- 

 trasted with the obdiplostemony of the Ericales) , alternate arrange- 

 ment of the floral members, adnation of the filaments to the corolla, 

 and dehiscence of the anthers by longitudinal slits, form weighty 

 objections against an assignment of the Lennoaceae to the Ericales. 

 Further, certain other characters possessed by the Lennoaceae, 

 viz., their short and solid style, normally developed endothecium, 

 separate pollen grains, and multilayered seed coat, render its 



10 See also Hagerup (1946) who has confirmed this opinion as the result of his 

 morphological studies on the Empetraceae. 



