166 MK. w. mitten's list or samoan mosses. 



conjecture that Gaslondia of the latter botanist was the same 

 genus having proved correct. Messrs. Brongniart and Gris show 

 that it agrees in all respects with Eugenia (sect. Jamlosd)^ except 

 that the flowers are 3-merous — a character which, I believe, in 

 the whole tribe occurs only in one or two species of Myrcia, If, 

 however, it prove constant, it may serve to maintain the genus, 

 although in close proximity to Eugenia. 



Tribe IV. Lecttkii)!;^. 



The three well-known subtribes Barringtoniea?, Lecythidca?, 

 and NapoleonesB, which we have thought might weU be grouped 

 into a tribe under the name of Lecythidese, require no comment 

 on the present occasion. "We have proposed no alteration in the 

 circumscription of the thirteen genera of which the tribe is com- 

 posed, the American ones having been well distinguished by Berg, 

 and the Asiatic ones by Blume and others ; I have already had 

 occasion to allude to the affinities of Napoleona in a note on 

 AsteranthoSy printed in the third vol. of our Journal; and the 

 curious anomalies in the stamens and staminodia of these two 

 genera will, I believe, be fully discussed in a paper prepared for 

 the Society by Dr. Masters. 



A List of the Musci collected by the Eev. Thomas Poavell in the 



Samoa or IS'avigator's Islands. By Willta]^ MprxEN, Esq., 

 A.L.S. 



(With two Plates.) 

 [E^ad March 7, 1867.] 



Mosses 



thered chiefly in the Island of Tutuila. Including a few incom- 

 plete specimeus, about 100 species have been obtained ; with only 

 two or three exceptions, aU are from the bark of trees or from 

 decayed wood. As might have been expected, the Samoan Musci 

 manifest a close correspondence with those inhabiting the xigi 

 lis lands, and with them approach generally most nearly to the forms 

 which appear especially to belong to the islands of the Indian 

 Archipelago. Conspicuous amongst Mr. Powell's collection are 

 two species of Spiridens, generally resembling the original o- 

 Rienwardfii of Java, but both more robust— and two species ot 

 Garovaglia, flflso closely corresponding with other species found in 

 Java and the Moluccas. Accompanying these, Mr. Powell has 

 found a now genus, in its leaves and habit very nearly resembling 



um 



