86 THE JOURNAL OF BOTANY 



shorter seta ; and here it is certainly at times difficult to decide for or 

 against their autonomy. L. codspitom Hedw. is one of these forms. 

 In some of these cases the characters are purely ones of degree, and 

 taxonomy becomes to a considerable extent a question of taste. 



In the considerable list of synonyms I give below I have only made 

 fresh reductions when I have been able to examine types or original 

 or well authenticated specimens. A further number, to judge from 

 the descriptions, especially of S. American species, will certainly have 

 to be added, but however convincing the description (or poverty of 

 description) may be, 1 have thought it better to omit all reference to 

 those of which descriptions alone have been available. I append a 

 few notes on some of the reduced species. 



I have, after some hesitation, desisted from any attempt to group 

 the various forms under varietal heads. The variations run on so 

 many lines, and are so little correlated together, that it is practically 

 impossible to classify the forms except on the basis of a single 

 character, such as the form of leaf, the acuteness or otherwise of 

 apex, and so on ; and none of these appear to afford a basis, for any- 

 thing but a purely artificial classification. One of the most distinct 

 forms is that represented by H. Dicnemonella (CM.), where plants 

 of a very small size are accompanied by an unusually obtuse, elliptical 

 leaf with remarkably wide cells and very small s])orophyte. Several 

 of the African species which I have reduced come under this head, 

 but I have also found the same leaf form in some South American 

 plants not alwaj'^s associated with the same fruiting characters. 



11. Kegdiaiium (C. M.), Surinam, Herb. Doz. & Molk., is the 

 ordinary R. cassjyitosiim, the form with wide homomallous leaves 

 and i-cither wide cells. 



IL Sduloiiia (C. M.), Dusen, No. 87, is a form showing transi- 

 tion to the '"'' Dicnemonella''' form just referred to. This transition 

 is still more clearly shown by R. suhcurvulum (C. M.) Broth., coll. 

 Zenker. R. afvo-demissuni (C. M.) is also certainl}^ the same thing. 

 C. Mueller, in fact, compares it with R. coespitosum, and is only 

 able to say for it that it is smaller with quite entire leaves, more 

 laxly areolate, and smaller capsule. None of these characters would 

 remove it from R. coespitosiun as understood in this paper. H. JBos- 

 loelli Geheeb, ex Herb. Bos well, he adds, is very near it. 



R. coespitans Schimp., Guadeloupe, I'Herminier, Herb. Schimp., 

 is a dense, robust, beautifully golden form, but structurally exhibits 

 no differences. 



R. aguatum (Hampe). This appears to be a very plastic plant. 

 I first examined the type in Herb. Hampe, coll. Lindig (without 

 number). This has rather narrow acuminate leaves, with elongate, 

 rather pellucid cells, and ^^erichsetial leaves very finely filiform- 

 subulate, like L. ccespitosa as figured by Hedwig. I concluded, 

 therefore, that it must be distinct. Subsequently, however, I 

 examined the specimen in Bescherelle's herbarium, No, 213o coll. 

 Lindig, and found that this was certainly only a small form of 

 R. coespitosum, with quite normal, widely-pointed perichaetial bracts. 

 I suspected a mixture, but on further examination I found that the 

 two insensibly intergrade, and while the bulk of Lindig's plants 



