EMBRYOLOGY IN RELATION TO TAXONOMY 359 



endosperm cells; persistence or gradual disappearance of endo- 

 sperm in the mature seed. 



11. Embryo. Relation of the proembryonal cells to the body 

 regions of the embryo ; form and organization of the mature embryo ; 

 presence or absence of suspensor haustoria. 



12. Certain abnormalities of development. Parthenogenesis; apog- 

 amy, adventive embryony, polyembryony etc. 



An evaluation of the characters mentioned above has been of 

 considerable service in the determination of the proper position of 

 several difficult groups and subgroups, and sometimes it has given 

 a new orientation to our ideas of their affinities. Without going 

 into details, for which a reference may be made to the work of 

 Mauritzon (1939), the following selection is offered as an illustration. 



Empetraceae. Don (1827), who first erected the group "Empe- 

 treae," considered it to be so different from the Ericaceae that he 

 rejected any possibility of a close alliance between them. He be- 

 lieved instead that the Empetreae was more closely related to the 

 Euphorbiaceae and the Celastraceae. "The Euphorbiaceae and 

 Empetreae agree in the imbricate aestivation of the calyx, in the 

 stamens being opposite to the divisions of the calyx, and both of 

 these being of an equal and definite number; in having bilocular 

 anthers ; in their superior ovarium ; in the plurality of styles ; in their 

 divided stigmas; and lastly in the arrangement of the ovula, and the 

 presence of a copious albumen." He went so far as to say that 

 whether the Empetreae was to be considered as a section of the 

 Euphorbiaceae or a separate family allied to the latter was a matter 

 of individual taste. 



Bentham and Hooker (1880) felt less sure about a relationship 

 between the Euphorbiaceae and Empetraceae and assigned the 

 latter to their or dines anomali under the Monochlamydeae. Shortly 

 afterwards, Pax (1896) also denied that the Empetraceae showed 

 any recognizable affinities with either the Ericaceae or the Euphor- 

 biaceae. The floral structure and in particular the structure of the 

 ovules left no doubt in his opinion that it was to be placed in the 

 order Sapindales close to the Celastraceae 1 and Buxaceae. 



However, Agardh (1858), Gray (1858), Baillon (1892), and Hallier 

 (1912) considered the Empetraceae to be related to the Ericaceae, 



1 Among recent writers Hutchinson (1948) still thinks that the Empetraceae has 

 its nearest relatives in the Celastrales. 



