Pammel — Anatomical Characters of Seeds of Leguminosae. 131 



researches were not confined to a single order but covered 

 Anonaceae, Berberidaceae, Bixaceae, Cahjcantliaceae, 

 Canellaceae^ Capparidaceae, Garyophyllaceae^ Cistaceae, 

 Cruciferae, Dilleniaceae, Franheniaceae, Fouqinerieae, 

 MagnoUaceae, Menispermaceae, Papaveraceae, Piitospora- 

 ceae, Polygalaceae, Portulacaceae, Banunculaceae, Reseda- 

 ceae, Sarraceniaceae, Tamaricaceae ^ Tremandraceae, and 

 Violaceae. Various characters are used, such as simple or 

 compound trichomes, presence or absence of crystals, stomata, 

 and fibro-vascular bundles. In a general comprehensive 

 paper on the subject (460. 461) his views have not changed 

 materially, and he adds that taxonomists often fail to obtain 

 all of the material in collections necessary to properly describe 

 a. species, hence the descriptions are imperfect. These uncer- 

 tainties will cease when anatomy occupies a place of merit in the 

 system of classification . This calls to mind the work of Solere- 

 •der (444), who shows that Masters wrongly determined a 

 Bragantia sent to him by Dr. Cleghorn. Solereder states that 

 its anatomical structure shows that it does not belong to the 

 order Aristolochiaceae but probably to the Menispermaceae. 

 In his exhaustive account of the anatomical characters of 

 this order he includes the structure of the seeds of many 

 species. 



The nature of the epidermis, stomata, secretion glands, 

 and sclerenchyma affords valuable diagnostic characters. 

 Eadlkofer (429), 1883, emphasized the importance of this 

 application of the anatomy of plants, in his address before the 

 Munich Academy. Pax (420), who has studied the order 

 Fitphorbiaceae, established an anatomical system. His 

 studies supported the divisions made by Miiller instead of the 

 Baillon, and Bentham and Hooker systems. Quite recently, 

 Van Tieghem (457) has made use of some anatomical charac- 

 ters in the special part of his work Elements de Botanique. 



The Leguminosae have been investigated by Jaensch (383), 

 Warburg (465), Schube (440), and Van Tieghem (455). 

 Jannicke (384) found eight types of stem in Papilionaceae. 

 It was not difficult to find characters of systematic value. In 

 Vicia and Tre/o?«<m, large genera, different characters occur; 

 these are, however, not always so marked as the gross morpho- 



