162 ON A NEW GENUS OF TURNERACE.E FROM RODRIGUEZ. 



floraceae, their main distinction being the absence of the corona. 

 Seemann, indeed, considering the petaline fringes in Erblichia 

 the homologue of a corona, united the families ; but this (as 

 pointed out by Bentham and Hooker) is a mistake, and the fringes 

 of Erblichia point rather in the direction of Pangiae, a tribe of 

 Bixineae ; and their habit, fugacious petals with no corona, suffi- 

 ciently distinguishes Turneraceae from Passifloraceze. Matliurina 

 does not in any way affect the relationship of the families ; for 

 I apprehend the glands on the sepals have no homology with the 

 corona in Passifloraceae. 



"With Bixineae a much closer affinity exists, as also indi- 

 cated by Bentham and Hooker ; and in fact Erblichia and 

 Matliurina so completely unite the families that I fail to 

 find any point of absolute distinction. I am therefore in- 

 clined to agree with Baillon (Hist, des Plantes, vol. iv. p. 293), 

 who makes Turneraceae a tribe of Bixineae, or at least to re- 

 legate the family to the vicinity of Bixineae rather than of Passi- 

 floraceae. Matliurina also establishes a very important connexion 

 between Turneracese and Samydaceae, the connexion bearing 

 upon geographical distribution. Samydaceae are a link between 



Passifloraceae and Bixineae — from the former differing in the 

 absence of a corona, and from the latter in the perigyny of the 



stamens. But the relationship with the latter is so close that 

 they are frequently united as one family. With Samydaceae, 

 then, Matliurina connects Turneraceae in many particulars; and 

 specially w T ould I notice that in the tribe Homaliese we find the 

 type of those glands (calycine) which are so characteristic a 

 feature in Matliurina ; and, further, one species of Homalium is 



- 



Mauritian, and two genera, Calantica and Bivinia, in which the 

 glands are well developed, are endemic in Madagascar. This close 

 affinity of Mathurina with Mascarene types, and the relationships 

 it establishes betwixt Turneraceae and families in which such 

 types are found, seem to me to render less surprising its occur- 

 rence in these islands — anew genus of a family hitherto unknown 

 in these islands, and with its nearest affinity a monotypic Central- 

 American genus. 



• 



