RirAPlirUOSTEOIUM CaCSP.TTO.SlIAI S3 



not to be includud in the " Fornienkreis " of E. ccesj)ifosiu//, but 

 it is at the far extreme of the forms ; if it is to be so considered there 

 would certainly be no need for the present paper, for all the characters 

 which have been held to separate the plants 1 am here uniting are of 

 minute importance as compared with these two extremes. Apart 

 from this consideration, then, C. Mueller's description of //. loxcuse 

 gives no distinctive charactei's as sej)arating it from K. coeHjiitosiiin 

 that will stand cross-examination. He describes the leaf -margin of 

 II. caespifosum as '* erecto vel vix rellexo," and that of II. loxense as 

 " valde revoluto," but any stem of the plants under consideration will 

 almost certainly show leaves with margins erect and others more or 

 less strongly recurved or refiexed side by side on the same branch. 

 The form of capsule indicates, it is true, some difference, which is 

 certainly manifested on the actual plants, but it is not correlated 

 with any other characters, and it depends to a considerable extent on 

 the degree of maturity of the fruit when gathered. The African 

 plants show a greater diversity in form of ca]3sule, while preserving 

 the other characters unimpaired. 



Mrs. Britton and Mr. R. S. Williams have given some study to 

 this group, and have a wide acquaintance with the South and Central 

 American forms of Hhaphidostecjium. Mrs. Britton writes, in answer 

 to my enquiries, that they have not had an opportunity of studying 

 the principal types, which are in Europe, but that they have given 

 some attention to the group and been puzzled by their variations ; she 

 has considered that R, loxense and R. gaU2)ense differ from R. 

 cuespifosum in having the leaves more secund and the pericha^tium 

 more serrate. She adds that Mr. Williams has not seen the West 

 Indian plant growing, but is inclined to recognize R. loxense as dis- 

 tinct, and relies on differences in the perichaetial leaves, and the 

 amount of serration of the leaves. 



It was clear that the only way to resolve the problem was to 

 examine the types. Fortunately, most of these are to be found at 

 Kew ; Hooker's //. loxense is there in good material ; there are two 

 branches of Swartz's H. coespifosum in Herb. Hooker; and Mrs. Britton 

 has transcribed for me some notes and sketches made by Dr. A. Leroy 

 Andrews from the type at Stockholm. There is also an original 

 specimen of H. litliophilum Hornsch., and a large series of plants 

 under various names from the West Indies and Tropical America. 



As regards the perichietial leaves, these in R. loxense andi?. cocspi- 

 fosufn types are absolutely identical. In neither are they denticulate; 

 nor have I found anything amounting to denticulation in nearly all 

 the forms I have examined ; at the most they show an occasional 

 obscure subdenticulation or sinuation. I have rarely indeed found the 

 perichietial bracts in any species of Rhapliidostegium so constant as 

 they are through the whole range of at least the more robust forms 

 of this series of plants ; only in one {R. replicatum Besch. from 

 Reunion) have I seen them at all markedly denticulate. Besch erelle 

 separates this species from R. Duisahoanum mainly on the ground of 

 the perichffitial bracts " fortement dentees" — in the diagnosis it is 

 simply " denticulata " — but this condition does not occur on the type 

 specimens in his herbarium, where they are often distinctly denticu- 

 late but no more, and often not that. 



g2 



