-87- 



dited it to North America. It is an aquatic form of S. recurvumP; the differences 

 are those one becomes familiar with in aquatic forms. The cortical cells how- 

 ever remain those of 5. recurvum, tending to confirm one's impression that 5. 

 cuspidatum can not likewise be regarded as a mere aquatic form of 5. recurvum. 



As to distribution, S. recurvum is widely spread over the three northern 

 continents, and either itself or closely related species occur also in South America 

 and Africa. In North America it is not to be found everywhere, but may be 

 looked for generally north of the limits of glaciation, extending further southward 

 in the eastern states to Florida and through the tier of Gulf states to Louisiana. 

 In the Pacific coast region Washington seems to constitute the southern limit; 

 in the Rocky Mountains, Colorado. 



Var. tenue KlinggrafT, 1872. This is the small leaved form of the species 

 and probably deserves taxonomic recognition as differing so markedly in aspect 

 from the more normal forms that it is likely to be confused with other species. 

 The intergrading of forms is however such that I doubt if any two persons could 

 agree just where to draw a dividing line in case of a large number of specimens, 

 in fact I am not convinced that one person would draw it twice in exactly the 

 same place. Russow recognized in the work already referred to that the sub- 

 species contained two distinguishable series of forms, those with leaves clearly 

 undulate (undulata), and those without this character (imbricata). Warnstorf 

 in his revision of the group Cuspidata}^ which it should be said owed a great 

 deal to Russow, characterized them more in detail as the forms Warnstorfii Jensen 

 and tenue (KHnggraff). I have in North American Flora (19 13) tried to secure 

 a definite line of separation and at the same time recognize what seems the most 

 natural taxonomic grouping by restoring the undulate leaved forms to the species 

 and keeping under the variety only those with imbricate leaves, retaining then 

 the varietal name which both has priority and was originally used in this sense. 



The unfortunate attempt to make a species out of this variety or of these 

 various forms has led to considerable disagreement as to specific name. Roll 

 insists that his hrevifolium has priority. The publication of it upon which he 

 bases his claim^^ suggests a nomen nudum, as Warnstorf asserts it does, how- 

 ever, refer back to a series of forms characterized in Flora LXIX, i84ff. 1886, 

 some of which at any rate probably belong within our variety, though the S. 

 cuspidatum var. hrevifolium Lindberg, 1880 included in it and apparently giving 

 it its name belongs according to Warnstorf to S. halticum. Warnstorf himself 

 used a specific name parvifolium (1900) applied varietally by Sendtner, while 

 Russow as we have seen published in 1890 Jensen's angustifolium as a subspecific 

 name and Scandinavian bryologists still use this name as specific. In his last 

 work Warnstorf has given up 6". parvifolium as a species, and including it as a 

 variety under S. amblyphyllum has felt obliged to put its specimens with tri- 



12 Cf. H. Lindberg, Lotos, LL 123. 1903; Loeske, Zur Morphologie und Systematik der Laub- 

 moose, 50. 1910. 



" Verb. d. bot. Ver. d. Provinz Brandenburg, XXXH, 220. 1891. 

 " Bot. Centralbl., XXXIX. 340. 1889. 

 IS Pflanzenreich, 51: 2i4f. 1911. 



