262 THE JOURNAL OP BOTANY 



the mixed forest of Polynesian type, which finds its southern- 

 most limit in New Zealand. -The incidence of this formation 

 there is determined chiefly hy the rainfall, therefore characteristic 

 of the extreme north of the North Island and the west coasts of 

 both the North and South Islands, where the latter is very heavy. 

 This formation was particularly interesting to me, as I had just 

 spent the three previous months (equivalent to the same season 

 of the year) in Viti Levu, Fiji. There, in the glorious mixed 

 forests of that favoured island, the Liverworts, which form an 

 important component of the floristic composition of this type of 

 forest, were also, as was the case about Auckland, at their best. 

 The affinity and similarity of types and habitat was most striking, 

 both in the cryptogams as well as in the phanerogams. 



In Hepatics, Schistochila aiiijendiculata was a very conspicuous 

 and general type, draping and hanging from dead wood and trunks 

 of trees ; the yellow-green thalli often three-quarters of an inch 

 broad and hanging down five or six inches, terminated by one or 

 two sporangiophores. Schistochila heterodonta Col. of similar 

 size and habit was also general in the forests of Viti Levu, but 

 not seen in fruit. Symphogyna hrevicaulis was partial to tree- 

 fern stems, as is Sytnphogyna vitiensis Jack et Steph. in Fiji. 



The most striking analogy, however, was to be seen in a patch 

 of Treuhia insignis in full fruit, spreading over stones and stumps 

 from bank to bank in a small shallow stream which fell into the 

 Nihotupu Kiver, identical in form and habitat with T. bracteata 

 Steph., so familiar to me in Viti Levu. (See Journ. Linn. Soc. 

 Bot. xxxix. 194, 1909.) 



Goebel (Flora, v. 96, p. 187, 1906) records this plant (in 

 sterile condition) in New Zealand at Dunedin and Otira Gorge, 

 in the South Island, and near Wellington, at Kaitoki, in the North 

 Island. He remarks that the sporadic occurrence of this genus 

 in Java in comparison with its wide distribution in New Zealand 

 leads one to infer that it is there outside the centre of its region 

 of distribution. One species having been found in Samoa, he con- 

 cludes the genus undoubtedly belongs to the Pacific-New Zealand 

 region, and that Java is one of its most western localities. The 

 fact that T. bracteata Steph., the Samoan species, was found by 

 me occurring abundantly in Viti Levu, up to 3000 ft., whereas 

 recently, in North Borneo, I did not see the genus at all, would 

 seem to support this hypothesis. 



Goebel observed a difference in the size of the New Zealand 

 plants which came under his observation, as compared with the 

 Javan species T. insignis, which was first collected by him. He 

 mentions the fact that the New Zealand plant is smaller ; at 

 Te Aroha I saw some thalli similar in habit and of smaller size, 

 but not being in fruit they were not collected. My Nihotupu 

 specimens of T. insignis, however, were very much larger and 

 identical in size and habit with T. bracteata, as seen by me in Fiji. 



In all, forty-one species in twenty-one genera were collected, 

 but these only represent those seen in fruiting condition. It was 

 my intention to have given the distribution in relation to the 



