342 OV A SECOXK COLLECTTOIT OF FBR1T8 FROM SAMOA. 



greatest similarity in PJaniago and Logo.nia, although I have not 

 tuiind the testa mucous in species of the latter genus. But the 

 mucilaginous testa is not even of generic value for discrimination in 

 other Orders, as researches on Eri/simum, Lepidium, Pi/rus, and other 

 genera have clearly demonstrated ; and this character is also only one 

 of degree among the various species of Plantago. The habit of Plan- 

 taginem, it must be admitted, is not repeated among any Loganiaceous 

 genera hitherto known, however great even the external dissimilarity 

 among Loganiacece themselves ; but outer appearances separate even 

 many species of Plantago widely from each other, while Littorella 

 stands as much by itself in habit as Limosella does among Scrophu- 

 larincB. The circumscission of the capsule, though universal in the 

 great typical genus Pkmtago, is neither in this or any other Order of 

 more than generic importance. Opposite leaves, necessary in true 

 Loganiacece, are rare in Plantaginccz, but do occur. The strong longi- 

 tudinal nerves of the leaves which many species of Plantago exhibit 

 remind us of those of Strychnos, and particularly of many plants of 

 the closely allied Gentianece. The placentation, however, affords a 

 s'rong mark of distinction between Loganiacece and Plantaginecn. The 

 placenta of Plantago, often free in the ripe fruit, gives grounds for 

 recognising a relationship to Priimdacece, the seeds of which are also 

 frequently peltate. The alliance to Plumhaginece, so frequently 

 insisted on, is obscure ; indeed, it appears tome that this family might 

 readily be transferred to near the Curvembryonatoi, which again form 

 tne best transit from Tlialamiflorm to CahjciflorcR, whether mono- or 

 dichlamydeous, the Plurabagineous plants being moreover mostly 

 dialypetalous. And finally, it may be added that really no difficulty 

 exists to dispose of all the real MonoclilamydecB ( Conifercc and Cgcadecs 

 considered achlamydeous) among the Thalamifiorcc and Calycijlorce, 

 especially since the discovery of the genus Buckley a seems to have 

 proved the floral envelope of Proteacecc and Santalacecu to be petaline 

 and analogous to that of Loranthacece. This view, furthermore, seems 

 confirmed or strengthened by the articulation of the petals of many 

 Proteacecc with the pedicel, the real calyx thus being obliterated, as 

 in Biplolccna and Asterolasia among RutacecB, not to allude to some 

 similar instances in several other Natural Orders. 

 Melbourne, I'^th July, 1876. 



ON A SECOND COLLECTION OF FERNS MADE IN SAMOA 



BY THE llEV. S. J. WHITMEE. 



By J. G. Baker, F.L.S. 



At page 9 of the present volume of the Journal I gave an account 

 of a collection of Ferns made in Samoa by the llev. S. J. Whitmee, of 

 the London Missionary Society. This summer he has sent a consider- 

 able supplement, the additional species in which are as follows. As 

 before, the numbers in braci<ets indicate the position of the novelties 

 in the series followed in Hooker and JJaker's Synopsis, and those not 

 in brackets are those under which Mr. Whitmee sends the plant, 



