96 



lu this section nearly all the species have a decidedly different aspect when 

 young as compared to mature plants, a fact which has prompted even such a 

 careful worker as Asa Gray to describe the yimni;- tlowerless plant of C. Icpto- 

 stegkt, his own species, as a variety pinnatiloba of C. coriacea an entirely dif- 

 ferent plant. The young plants of ('. acuhatiflora and C. iviioiiuuttlia have 

 deeply bipinnatisect leaves, usually iirickly but otherwise resembling a deeply 

 divided fern-frond; old plants however have usually entire leaves. 



As has been remarked, the writer believes this section to be the oldest ; first, 

 because it is found on all the islands of the lii'oup with the exception of Niihau 

 and Kahoolawe; their nonexistence there is due to the .smallness of the islands in 

 ((Uestion and consequently do not otter the range of conditions found on the 

 larger islands, so essential to their development. 



While in Berlin the writer had occasion to examine species of both genera. 

 dentropogon and Sipliocanipyliis. as far as represented in the Royal Botanical 

 Museum, and b\- studying them came to the above conclusion based on the fol- 

 lowing facts: 



Outside of some of our Hawaiian genera of Lobelioideae, Cciil nijiiiijini is the 

 only genus with true baccate fruits; the species of Centropogoii. like Cyaiira, 

 are nearly all shrubby; only a few ai-e herbaceous. The fruits of the genus 

 Pratia have a fleshy pei'icarp; the species are however in nearly every instance 

 prostrate herbs and have no resemblance to our jirborescent Hawaiian forms. 

 SipJiocampylus. a South American genus Avitli over 100 species has capsuhir 

 fruits opening at the apex into two valves. 



The only important dift'erenee between Ci ii/nipngd)! and Cgnui <i is the corolla, 

 which is very shortly slil at tlie baek'. in the former, while the Hawaiian species 

 of ('[jaiua are deeply slit to beyon<l the middle. The calycine lobes are also a 

 very variable character in Ccntropogan, ranging from linear-subulate, elongate 

 to linear lanceolate, or are reduced to teeth; the tlowers are arranged singly in 

 the axils of the leaves, a character never occurring in Cyanca, or are racemose 

 or in terminal corymbs The genus Apetaliia. with its single species (.1. raia- 

 teensis) from the islands of Raiatea and Tahiti, seems to be an intermediate 

 between the strictly capsular and baccate genera. Apetaliia possesses an indehi- 

 scent ( ?) rather dry fruit which is unilocular, and possesses an undivided 

 stigma. S(I( rollii ca with two species in the Society Islaiul (Tahiti), and one in 

 Rarotonga, is of great interest. The fruits of this genus are capsular and the 

 seeds are dispersed by two pores at the apex of the capsule. 



