264 THE JOUllNAL OF UOTANY 



similar forms of S. imiiulatum still persist. The Acutifolia rarely 

 compete for the occupancy of deep moorland pools, but otherwise 

 occupy every kind of habitat from wet bogs to damp peaty patches 

 by roadsides. Plants of the C i/mh if oli a gvouy^ avoid both tlie drier 

 habitats and deep pools, being usually plants of moderately wet moors 

 and bogs. S. jyttpiHosum in West Cornwall is generally associated 

 with Cotton-grass, Molinia coendea, Hypericum elodes, Pedicularis 

 palustris, and plants of similar requirement as regards moisture, 

 though the dense low tufts of forma conferta may intrude into 

 sli«'htly drier ground. It follows that as lands are reclaimed or 

 become drier through natural causes the first plants to disappear are 

 the Gu^pidafa, with the more robust Stcbseciinda (though these may 

 persist in springs and deep ditches), and next the Cymhifolia ; while 

 the most persistent are the more delicate forms of the Suhsecioida 

 and some of the Acutifolia. In fact, wherever a few thin tufts of 

 Sphagna remain by damp roadsides or in tield-borders they prove to 

 be either S. plumiilositm (or one of its near allies) or delicate forms 

 of >S'. snhseciuidnm, S. aitriculatum, or ^S*. inundatum, and where, as 

 sometimes happens, peaty ground merges gradually into pasture, 

 these plants occupy the outmost ranks. Thus (to quote a typical 

 case) in Lambourne Valley, about a mile and a half in length, 

 Spihagna remain only in a few square yards of peaty ground by the 

 stream in the border of a field, associated with fruiting Hj/locomiuni 

 squarrosum and partly shaded by furze bushes, and the actual plants 

 occurring are S. plumulositm var. viride and var. versicolor f. tenel- 

 lutn, and S. inundatum var. diver sifoUum f. etirycladum. 



It is noticeable that the plants of drier habitat among the Suh- 

 secunda are often of somewhat plumose appearance and have the 

 margins of the branch leaves more or less incurved, but do not 

 exhibit torsion of the branches ; this in my expei'ience is restricted to 

 plants of very wet ground and is often most apparent in the upper 

 (aerial) parts of such plants as grow in shallow water. 



Another fact which may have some meaning is that these plants 

 of drier habitat all belong to the sub-group in which the outer 

 (dorsal) surface of tlie branch-leaves is abundantly supplied with 

 pores, these pores being most dense towards the point of the leaf — 

 the part most exposed to the atmosphere — whereas the plants with 

 fewer pores, or with pores differently situated, grow submerged or 

 with only the growing point exposed. Plants of the first sub-group 

 grow, of course, in very wet places, but not, I believe, floating or 

 submero-ed. All my Cornish gatherings of Suhsecunda fall readily 

 into one or other of two groups : — 



{a) Plants of aerial growth, /. e. growing on comparatively finn 

 ground, or if in shallow water then with erect stems rising con- 

 siderably above the surface. All are plants with dorsal pore develop- 

 ment, viz.:^ — S. suhsecundum, S. inundafum, S. auriculafum, S. 

 aquatile,SLudi S.eonfortum. 



(b) Plants of aquatic habit, floating or submerged: — S. ohesum, 

 S. Camus a, S. crassicladum, and S. turgiditlum. 



I do not know to what extent submerged forms of the first sub- 

 group ma v occur in other localities, but if it be considered that the 



