-4— 



of these cells and that next below it in vertical series is not a plane 

 one, but projects saccately, or in extreme cases in funnel form into 

 the cell next below, providing connection with it by a pore at the 

 end of the funnel. 



These characters seem to distinguish the form amply from its 

 parent species, 5. imbricatum, but it is not inconceivable that further 

 study of it in the field may reveal more nearly intergrading forms. 

 Its range falls essentially within that of 5. imbricatum and that in a 

 characteristic region, which has more than all others in orth 

 America evolved secondary forms from species of more general dis- 

 tribunon in the northern hemisphere. I have seen specimens of it 

 from New Jersey, Georgia, Florida and the West Indian island of, 

 Guadeloupe, which with the original Portorican station represent its 

 present known range. It is in New Jersey aquatic in its preferences,. 

 being found according to Austin in shallow ponds. Further observa- 

 tions and collections are very desirable, as its reproductive organs 

 and fruit are entirely unknown. 



2. Sphagnum imbricatum Hornschuch, 1865. Russow's publication 

 of Hornschuch's herbarium name of this species is generally re- 

 garded as adequate, though it contains no full description, hardly 

 more in fact than mention of a single characteristic feature, but that 

 of such a sort that there is no mistaking the plant's specific identity. 

 Under the previous species we have already characterized it in the 

 main ; where the fringe-fibrils are present they are a sufficient diag- 

 nostic character, except in so far as it might be confused with 5. por- 

 toricense ; it must be noted, however, that they are not necessarily 

 present throughout the leaf, but may be lacking in a greater or less 

 portion from the apex downward, in some specimens occurring only 

 in a few cells of the middle basal part and in the variety lacking en- 

 tirely. There is in this feature constancy only in that in a given tuft 

 of plants one finds the branch leaves of the various plants very 

 closely similar, but in a number of collections all degrees may be 

 represented, so that a division at any point becomes an arbitrary one. 



An interesting character of this species was observed by Russow^~\ 

 but his description of it seems to me not altogether accurate. It has to 

 do with the fibrils in the cortical cells of stem and branches. If one ex- 

 amine the inner walls of these cells where they are in contact with the 

 wood cylinder, one finds a very notable and, so far as I have observed, 

 entirely constant difference in that S. imbricatum (with 5. portoricense) 

 shows here not the normal continuation of the fibril-bands of the rest 



(1) Russow, Beitrage zur Kenntnis der Torf moose, 21. Dorpat, 1865. 



(2) Zur Kenntnis der Subsecundum und Cymbifoliunnigruppe der Torfmoose, 

 74, 93, 95. 1894. 



