of the Natural Ordir of ' Myrti 277 



cxpofurc to the air. I have confulted his fpecimens, and find no 

 realbn to doubt their being the fame as mine. 



7. M« annularis, foliis fparfis linearibus mucronatis apice re- 



curvis, floribus lateralibus, filamentis longifhmis linearibus 

 apice radiato-multifidis. 

 Metrofideros armilJaris. Gcertn. San. v. 1. 171./. 34- f. 5. 



This has much the habit of a Diofma, in the leaves efpecialIy T 

 which, in a garden fpecimen with which I was favoured from Mr. 

 Robertfon's at Stockwell, are very diilinclly marked with a row of 

 refinous fpots on each fide the mid-rib at the back, but thele are 

 lefs vifible in the wild plant. The flowers are white, cluttered 

 about the lower part of the branches, in the form of a long fpikc. 

 The footftalk or claw of the united filaments is very long before it 

 branches off, even thrice the length of the petals. 



8. M. gehiftrfoPa, foliis fparfis lanceolatis mucronatis trinerviis 



multipunclatis, ramulis floriferis terminalibus Iaxis, fila- 

 mentis apice radiato-multifidis. 



Sent from Port Jackfon by Mr. David Burton. It is in fome 

 refpe&i like M. mdoja, but the leaves are lanceolate rather than 

 linear, not above half fo long as in that fpecies, nor fo rigid and 

 pungent. The branches terminate in loofe {pikes, from the top of 

 which the branch is at length continued, as in the other fpecies. 

 The flowers are ieflile, in alternate pairs, white. Claw of the 

 (lamina twice as long as the petals before the filaments branch off. 



* * 



Fofiii 



