Heller : PLANTS OF THE HAWAIIAN ISLANDS. 857 



vex, keeled, slightly ciliate; corolla three-eighths of an inch 

 long, streaked with purple; petals connivent for nearly their 

 entire length, forming a tube, oblong, blunt, the ends some- 

 what dilated, twisted and recurved; stamens five, short, barely 

 exceeding the ovary, filaments flat, slightly expanded above, 

 anthers broad, with short, curved appendages at the base; 

 style the length of the corolla tube, almost straight, slightly 

 thickened at the apex, stigma minute. 



The type is No. 2828, collected above Waimea, Kauai, at an 

 elevation of 2000 feet. I was growing in a ravine between the 

 forks of the Waimea river. Hillebrand seems to have had 

 specimens of this species, collected by Knudsen, probably on 

 the west side of the Waimea river. He referred them, how- 

 ever, to Isodendrion longifolium, specimens of which I have seen 

 in the herbarium of Columbia University, and which is a very 

 different plant, as evinced by the specimens mentioned above, 

 and by the plate, in the atlas of the Botany of the U. S. Ex- 

 ploring Expedition. Plate LIV shows a flowering stem, and a 

 flower and stamen enlarged. 



YIOLA L. Sp. PI. 933. 1753. 



Viola chamissoniana Gingins, Linnaea, 1:408. 1826. 



Hillebrand records a variety of this species from Kauai, 

 which has "young shoots and inflorescence puberulous." In 

 specimens collected along the edge of the plateau above Wai- 

 mea, the leaves are pubescent underneath, even the older ones. 

 It is a spreading bush, three or four feet high, with pale violet 

 flowers. The type was collected by Chamisso, on Oahu, in 

 1807. 



October 12 (2880). 



Viola kanaensis A. Gray, Bot. U. S. Expl. Exped. 15:85. 

 1854. 

 These specimens differ considerably from Gray's original 

 description, who, it seems, did not have the early, large flow- 

 ered form. He says : "The flowers of the specimens are prob- 

 ably late ones, with the petals smaller than when fully de- 

 veloped, as they are not quite so long as the calyx. * * * 

 What strengthens the suspicion that these are only such pre- 

 cociously fertilized and cryptopetalous flowers as produced by 

 many violets, is that the stamens are scarcely shorter than the 

 petals.*' The description calls for "peduncles IV to 2 inches 

 long, nearly the length of the petioles, glabrous, furnished 



