THE PTERIDOSPERMS AND THE SEED 259 



theory of the derivation of an Angiosperm flower from 

 an inflorescence of unisexual flowers, Lotsy remarks that 

 the author bases his view that the primitive Angiosperm 

 flower was unisexual on the supposed origin of such 

 Angiosperms from the Gymnosperms, where herma- 

 phrodite flowers do not occur. The so-called " Mono- 

 chlamydeae " are thus to be regarded as primitive, 

 and amongst these the anemophilous forms stand lowest 

 of all. The Monochlamydeae undoubtedly possess very 

 simple flowers, but is simplicity, asks Lotsy, an invariable 

 indication of primitiveness ? May it not arise, secondarily, 

 by reduction ? May not anemophilous flowers be derived 

 from entomophilous ? Lotsy quotes at length, and with 

 approbation, from a paper by Robertson, published in 

 1904, on " The Structure of the Flowers and the Mode of 

 Pollination of the Primitive Angiosperms " in which the 

 author attempts to demonstrate (and, as it appears to 

 me, with success) that the primitive Angiosperms were 

 entomophilous, and that the Monochlamydeae are not 

 primitive but greatly reduced types. 



The leading exponent of the newer view is undoubtedly 

 Hallier, whose phylogeny Lotsy, in the main, follows. 

 Hallier's views which have been, strange to say, ignored 

 by Engler are as follows. I give you them in his own 

 words quoted from an article on " The Natural (Phylo- 

 genetic) System of Flowering Plants " published in 1905. 

 X ' ' The Angiospermae are a natural (monophyletic) group, 

 and not a polyphyletic one, as suggested by Engler in 

 Engler and Prantl's Nattir lichen Pflanzenfamilien. The 

 Amentaceae are not to be considered as old types, re- 

 maining in a lower state of development, and allies or 

 descendants of Gymnospermae, but, on the contrary, 

 as the highest and most reduced types of one of the lines 

 of Dicotyledons. They and all the other lines of Dicoty- 

 ledons have been developed by reduction in flower and 

 fruit from the Polycarpicae, the latter group being 

 derived immediately from Bennettitaceae or other 



17 A 



