BOTANY OF FERNANDO NOEONHA, 65 



et ferine -^^ unciam latse. Spiculse ad 20-florse angustse. Rhacliis 

 sinuata tenuis. Squamse lanceolatae subacutse, flavse ; carina 

 viridi. Stamen nnicum. Stylus bifidus ruber, squamam vix 

 superans. Caryopsis oblonga, apice breviter mucronato, testacea 

 puncticulata obscure biconvexa. 



This little Cyperiis was only met witb in clefts of rock on the 

 Peak, and on the slopes of Morro branco. It is 2-styled, the 

 nut being rather long and almost terete, showing very slight 

 traces of biconvexity. The spikelets, which are slender and have 

 the flowers rather distant, are curiously curled at the apex in 

 most cases. 



Ctperus compeessus, Presl, Bel. Scenic, iii. p. 177 ; Botth. 

 Gram. p. 28, t. 9. fig. 3 ; Boech. Gyp. Kerb. Berol. p. 121. 



Common on the main island, on the sea-shore in Peak Bay, and 

 also among the stones in the roads near San Antonio and in the 

 village, as far west as Leao Bay, where a very dwarf form with 

 short erect spikelets was seen. The plants were all small, as if 

 the species had only begun to establish itself in the island. 



Distribution. All tropical countries. 



C. viALis, n. sp. 



Annuus ; basi vaginis rufescentibus tecti. Folia linearia 

 acuminata acuta, 5 uncias longa, 1 lineam lata. Culmi validuli 

 triquetri 7-unciales. TJmbella triuncialis 4-6 radiorum erectorum, 

 spiculis pluribus in apicibus congestis, et nonnullis sessilibus in 

 medio umbellse. Bractese 4, umbella subsequales, lineares acu- 

 minatae. Spiculse 1|-1| uncias longae, graciles, 40-50-floraD, 

 colore G. coonpressi. Squamse lanceolatse mucronatse, margini- 

 bus late scariosis, carina viridi, lateribus ssepe rufescentibus. 

 Stamina 3, filamentis gracilibus. Antherse lineares apiculatse. 

 Stigmata 3, brevia rufa. Caryopsis oblonga angusta, trigona, 

 angulis obtusis, testacea. Ehachis gracilis exalata. 



Only two specimens of this plant were obtained, both at consi- 

 derable distances apart along the roadsides, in the central district. 



Prom its habitat I felt convinced it was an introduced plant, 

 but have never been able to match it with any species. It is 

 allied to G. rotundus, but the exceedingly long spikelets and 

 form of the nut make it quite different. It has, too, the green 

 and white colouring of G. compressus, and not the red glumes of 

 G. rotundus. Mr. C. B. Clarke, to whom I showed the plants, 

 pointed out the relationship to G. rotundus forma viridis, from 

 which, however, he considers it distinct. 



LINN. JOUEN. — BOTANT, VOL. XXVII. F 



