Platk 172.— PIMELEA LONGIFOLIA. 



Family THYMELJiACE^.] [Genus PIMELEA, Banks & Sol. 



Pimelea longifolia, Banks <& Sol. ex WiLsU. in Vet. Akad. Handl. Stockh. (1818) 280- 

 Cheesein. Man. N.Z. Fl. 609. 



Of the twelve species of Pimelea found in New Zealand this is by far the 

 handsomest and the most attractive, and it says little for the enterprise of colonial 

 horticulturists that it is so rarely seen in cultivation. It was first gathered by Sir 

 Joseph Banks and Dr. Solander at Tolaga Bay during Cook's first voyage, and was 

 subsequently noticed by them at Mercury Bay and Admiralty Bay. Curiously 

 enough, It does not seem to have been collected either by Forster in Cook's second 

 voyage, or by D'Urville or Allan Cunningham ; in fact, it was not until Mr. Colenso 

 commenced his botanical explorations that it was again met with. Subsequent 

 research has proved that although it cannot be called a common plant it is 

 neveTtheless thinly spread over the greater part of the North Island to the 

 south of the Bay of Islands, ranging from sea-level to quite 3.000 ft. It crosses 

 Cook Strait, and occurs in several localities on the west coast of the South Island 

 from Collingwood to Westport and Charleston, which is the most southern localitv 

 at present known. Mr. Bentham refers a plant from Lord Howe Island to the same 

 species, but I have had no opportunity of examining specimens from thence. 



Pimelea longifolia usually forms an erect branching shrub 3 ft. to 5 ft. or 6 ft 

 high and is easily distinguished from all the other species found in New Zealand 

 by the glabrous habit, large leaves, and many-flowered heads of large white flowers 

 Its nearest ally is undoubtedly P. Gnidia, which differs mainly in the much smaller 

 and frequently keeled leaves, and in the smaller flowers. The variety pulchella is 

 almost exactly intermediate between the two species, and might with perfect fairness 

 be referred to either. 



The flowers of P. longifolia are polygamo-dioecious, or possibly trimorphic 

 and should be carefully studied by any botanist who is fortunate enough to reside 

 in a locality where the species is abundant. I have figured three forms in the 

 accompanying plate, as follows : First (see fig. 1), what I take to be a hermaphrodite 

 flower, in which the perianth is large and broad, and the stamens and style are 

 about the same length, and well exserted beyond the flower. Secondly (fig. 4) a 

 male flower, with a much narrower perianth, and with the stamens agreeing with 

 the previous form m length, but with the style much shorter and entirely included 

 withm the perianth-tube. Third (fig. 5), a female flower, also with a narrower 

 perianth than in fig. I. but with the style conspicuously exserted, and furnished 

 with a large papillose stigma. The stamens have short filaments placed at the top 

 of the perianth-tube, and the small anthers are usually devoid of pollen. It would 

 be interesting to know whether the pollen of the hermaphrodite and male flowers 

 IS equally efficacious in fertilization, and also whether perfect fruit is produced 

 by the so-called hermaphrodite flowers, the stigma of which is certainly not so well 

 developed as in the female flowers. 



Plate 172 Pimeku loiu/i folia, drawn from specimens gathered in the vicmitv of Vucklaiid 

 b ig 1 hermaphrodite flower ( x 3) ; 2 and 3, front and back view of anthers ( x 6) ■ " 4 male flower' 

 with short style ( x 3) ; 5, female flower ( x 3) ; 6, longitudinal section of ovarv ( x 6) ; 7. seed (enlarged) ' 



