408 Transactions. — Botany. 



would be possible to take the opposite view, that such branch- 

 ing is a nascent feature, that it is a new feature in the 

 phylogeny. Bertrand regarded Phylloglossum as a form re- 

 duced on account of its semi-aquatic mode of life. But it is 

 necessary to point out that Phylloglossum is not a semi- 

 aquatic ; Bertrand never had the advantage of seeing the 

 plant in its native home. Phylloglossum, it is true, being a 

 very small plant, can only grow whilst the surface soil is 

 fairly moist, hence it forms a tuber and rests during the dry 

 season. So far as I have seen, the plant grows rather better 

 on a hill-top ; or. at any rate, it grows there at least as well 

 as it does lower down on the slope, and I have never found it 

 in an actual swamp. It grows well on a slope where water 

 can never lodge. Its roots spread rather horizontally, and 

 seldom far downwards in the ground, as though it objected to 

 a waterlogged soil. 



Whilst it is possible that evidence may yet be adduced 

 that Phylloglossum in some measure owes its simplicity to 

 reduction, there appears to be little evidence for this at pre- 

 sent. On the other hand, it may yet prove that Phyllo- 

 glossum is an exceedingly primitive plant, possibly the most 

 primitive of existing pteridophytes. We have an explanation 

 ready to hand of this exceptional retention of ancient charac- 

 ters — namely, the annual renewal of the embryonic stage in 

 the formation of the protocorm. But. however this may be 

 decided, the relatively simple character of the gametophyte 

 and the comparison of the mature sporophyte with the em- 

 bryo of Lycopodium eernuum are in favour of the view that 

 Phylloglossum is the most primitive of existing Lycopodina. 



