CLASSIFICATION OF ANGIOSPERMS 147 



L'origine et Ic systeme phyUtique des Angiospermes exposes a 

 Vaide de lew arbre genealogique. Archives Neerlandaises 

 des Sciences Exactes et Naturelles : Ser. III. B i. (1912), 

 pp. 146-234. Seven figs, in text, and six phyletic tables. 

 By Hans Hallier. 



This ambitious and, in many ways, remarkable essay repre- 

 sents, according to the author, his fifth attempt at a natural 

 system of classification of the Angiosperms. In the brief intro- 

 duction he claims that he has succeeded in defining sharply 

 the limits of the various orders, and not unjustly criticises " ce 

 qu'on appelle le systeme d'ENGLER " (the italics are ours) as 

 marred by more than one " agglomerat tout-a-fait heterogene." 



The introduction, of about three pages, is followed by a 

 lengthy " abrege raisonne de ce systeme " (pp. 148-202). A 

 list of twenty-three relative papers then precedes the exhibition 

 of the system itself, which occupies the remainder of the work. 



The essay reveals throughout the remarkable grasp of details 

 and the familiarity with all the groups of Angiosperms which the 

 writer undoubtedly possesses. He is thoroughly at home with 

 his subject ; but he affords us but a poor view of his household 

 gods. He has, it would appear, endeavoured to compress into 

 ninety pages what could hardly be made clear in five times 

 that number; the result is that, although he makes us feel 

 that lie is sincerely convinced, he leaves us unconvinced : and 

 this the more for not a few drastic alterations upon recognised 

 systems which he urges upon our acceptance. Thus, to quote a 

 few instances at random, Crassulaceae are associated with the 

 families comprised in Engler's Centrospermae ; Chloranthaceae 

 stand next to Lauraceae : Buxus and Batis are included in 

 Hamamelidaceae ; Boraginaceae, including Lennoaceae, are sepa- 

 rated widely from their customary associates and appear close 

 to Loasaceae, CampanulaceaB, and Compositae ; " Amentaceae," 

 Urticaceae, and Aceraceae stand in close connection with 

 Terebinthaceae ; Engler's Ebenales are entirely disintegrated — 

 Symplocaceae following close upon Celastraceae, Ebenaceae and 

 Styracaceaa are included in Santalales, while Sapotaceae, regarded 

 as the progeny of Linaceae, are assigned separate ordinal rank ; 

 Euphorbiaceas are classed with Salicaceaa and Elacourtiaceae 

 in " Passionales ; " Salvadoraceae are reckoned among the near 

 relations of Polygalaceae ; Primulaceaa and Myrsinaceae are 

 referred to an Ochnaceous ancestry and removed far from 

 Centrospermae — though Plumbaginaceae are associated with the 

 last-named group. 



At the outset of his exposition the author supports vigorously 

 the primitiveness of the Eanalian floral type (Polycarpicees) ; 

 and from this type he proceeds to derive all the Angiosperms 

 without exception. The Angiosperms are thus regarded as 

 monophyletic in the strictest sense of the term — i.e., they are 

 referable to a single angiospermous ancestor. This is the weakest 

 point of Hallier's system; for he neglects the much more probable 

 and reasonable view that certain groups of Engler's Archi- 



