390 Transactions. — Botany. 



other times dichotomous and fastigiate, in the axils and at the 

 nodes of which the sporangia are formed. There is no involucre, 

 but the sporangia are generally supported on a pedicel cell, and 

 are at first solitary, though at length more are evolved at the 

 side of the first. The lower cells of the branches are as 'much as 

 twelve times as long as broad, the upper are about six times as 

 long as broad. As in other plants of the genus, the cells when 

 put into fresh water burst with a crackling sound and dis- 

 charge their red colouring-matter. Cystocarps and antheridia 

 unknown. 



Distribution. — Australia, Tasmania ; New Zealand — River- 

 ton, The Nuggets, Port Chalmers, Moeraki, Oamaru, Lyttelton 

 (R. M. L.) ; Green Island Beach, St. Clair, Warrington (J. C. S.) : 

 Chatham Islands (Canterbury Museum). 



2. Griffithsia antarctica, Harv.. Fl. Ant., vol. ii., p. 488: J. Ag., 

 Epicr. Florid., p. 68. Plate XXV., fig. 2. 



Fronds tufted, flaccid, rather sparingly branched, dichoto- 

 mous, sometimes more or less irregularly pinnate towards the 

 upper part ; shorter, stouter, and much less tufted than G. 

 sonderiana ; from 8-12 cm. long, and about jmm. in diameter. 

 Cells of the main axis 1200-1500 p. long and 400-500^ broad. 

 The fertile branches are somewhat inflated, and at length give 

 off a short lateral branch, at the penultimate or antepenultimate 

 cell of which a cuplike involucre is formed enclosing the tetra- 

 spores. The cells beyond this gradually become absorbed or 

 disappear. The involucre consists of about 15 linear-oblong 

 ramuli, enclosed at first in a corolla-like cup, but afterwards 

 opening for the escape of the fasciculated tetraspores. In ma- 

 ture specimens the involucral cup frequently appears terminal 

 on a single obconic cell, but at other times it is supported by a 

 pedicel consisting of several cells. I have not been able to 

 determine whether this difference is specific or only varietal. 

 Antheridia and cystocarps unknown. 



Distribution. — Tasmania ; New Zealand — Akatore, Timaru, 

 Taylor's Mistake (R. M. L.) ; Green Island, St. Clair, Warrington 

 (J. C. S.) ; Foveaux Strait, Stewart Island (Lyall) ; East Coast 

 (Colenso). 



This plant may be readily distinguished from the preceding 

 species by the very different arrangement of the tetraspores. Its 

 generic position is uncertain. It may possibly belong to Pan- 

 dorea or Bornetia, but until the cystocarps are discovered it 

 should perhaps be left here. In Pandorea the ramuli surrounding 

 the tetraspores cohere as in a gamopetalous corolla. In G. 

 antarctica, according to Agardh, they are free, and this consti- 

 tutes one of his chief diagnostic distinctions between the two 



