PRESIDENTIAL ADDRESS. 45 



edonae is quite unnatural. In Mueller's system of the 

 Census some families are grouped more naturally, but he 

 left the Gymnospermse in their unnatural position, and he 

 succeeded only in putting a new artificial system in the 

 place of the old and very popular one. Alexander Braun's 

 system, published in 1864, is the first system that has a 

 real claim to the name "Natural System.'' A. W. Eichler's 

 came next ; A. Engler's, the system followed in the New 

 South Wales Census about to be published, is the latest 

 development. Of course no system can claim to be perfect. 

 If all vegetable forms that lived in former periods of the 

 earth had been preserved we would have a sure guide to a 

 natural system, but how few and imperfect are the frag- 

 ments preserved ! Some isolated forms can be only placed 

 tentatively at present, until later discoveries may show 

 their true affinities. 



Most important changes, revolutionising the whole nomen- 

 clature are made in the ferns. The system used in Hooker 

 and Baker's Synopsis Filicum is based, within the Tribes, 

 almost exclusively on the position and shape of the sori, 

 and the position and shape or the absence of the indusium, 

 and was followed generally by pteridologists, but it is 

 almost as artificial as Linne's system, based on the stamens 

 only. I will give here a few instances to what absurdities 

 unyielding adherence by the old system to one character 

 lead. 



Aspielium ramosum and Polijpodium tenellum are now 

 called Arthropteris obliterate!, and Arthropteris tenella; 

 the two Australian species and a third non-Australian one 

 form the well-defined small genus Arthropteris, allied to 

 Nephrolepis, but indistinguished from it by the leaves 

 being articulate on the rhizome. The two species have 

 the same venation, the same habit and have all characters 

 in common, except one has an indusium and the other not, 



