found on the mossy surface of rocks and logs under the shelter of Lepto- 

 sjjeniuim and Olearia Colensoi, its colour harmonizing so closely with that of 

 the moss that it is easily overlooked. It blooms in November and December, 

 and when fully mature the flowers have a purplish tinge. It should be 

 mentioned that the radical leaves are usually produced on special caudicles, 

 and not at the base of the flowering-stem. 



In the Manual I regarded Toumsonia as a close ally of Adenochihis, 

 relying principally on the structure of the column and the close similarity in 

 habit. Dr. R. Schlechter, who has recently published some interesting notes 

 on the genus (Fcdde, Hepertorium, ix, 249), suggests that it should be placed 

 in the vicinity of Aclanthus, with which it agrees in the smooth undivided 

 lip, devoid of any projections or protuberances beyond two or three obsisure 

 ridges. He also points out that the Tasmanian Acianthus viridis, Hook, f., 

 is very closely allied to T. deflexa, and must be placed in the same genus. 

 Townsonia is therefore a genus of two species — one confined to the Soutli 

 Island of New Zealand, the other endemic in Tasmania. Dr. Schlechter also 

 traces an affinity to the genus Stigmatodactylus, which has three species, found 

 respectively in Japan, India, and Java. 



Dr. Schlechter considers Townsonia to be an Antarctic type {typisch 

 antarktische). But surely he uses the term in an entirely different sense from 

 that understood by most New Zealand botanists. Genera like Colobanthus, 

 Acama, Azorella, Nertera, PliyUachne, Rostkovia, Oreobolus, &c., which have 

 species in the extreme south of South America, New Zealand, Tasmania, and 

 the circumpolar islands, may well be called Antarctic, and we are entitled to 

 speculate on their probable previous existence on the Antarctic Continent. 

 But in the case of Townsonia, whose nearest allies are the genera Acianthus, 

 Adenochihis, and Stlgmatodactylus, which have a purely Australian, Mela- 

 nesian, and Malayan distribution, the term seems inappropriate. And 

 especially is it so when we consider that the Orchidacece of New Zealand show 

 absolutely no relationship to the few members of the family found in the 

 extreme south of South America. 



Plate 198a. Chiloglotlis cornuki, drawn from specimens collected at Maungatapere, Whangarei, 

 by Mr. H. Carse. Fig. 1, side view of flower (x 3) ; 2, lip, showing glands (x 5) ; 3, front view of 

 column (x 5) ; 4, the same with the wings spread open (x 5). 



Plate 198b. Townsonia deflexa, drawn from specimens collected in the vicinity of Westport 

 by Mr. W. Townson. Fig. 5, side view of flower (x 3) ; 6, front view of same (x 3) ; 7, petal (x 5) ; 

 8, lip (the ridges much too conspicuous) (x5); 9 and 10, front and side views of column (x 5) ; 

 11, dehisced anther (x 6). 



