82 Rhodora [May 



form of vegetative reproduction, comparable with the gemmae found 

 in other genera of the Hepaticae. It is admitted that the presence 

 or absence of gemmae in certain cases affords a convenient method 

 for distinguishing between closely related species. The gemmi- 

 parous habit of Lophozia heterocolpa (Thed.) M. A. Howe, for example, 

 makes it possible to separate this species at a glance from L. Muelleri 

 (Nees) Dumort., in which gemmae are unknown. If this reasoning 

 is applied to caducous leaves (Bruch- or Brutblatter of Correns) 

 their presence, if supported by morphological features of even a 

 slight character, might well be made the basis for the segregation of 

 plants showing this feature from a species in which the leaves are 

 clearly persistent. In Schiff ner's forma densior of " Bazzania triangu- 

 laris Pears." and in Miiller's var. implexum of Pleuroschisma tri- 

 crenatum these conditions are apparently realized. The forms 

 designated by these names are clearly identical and differ from 

 typical B. tricrenata not only in having caducous leaves but also in 

 certain features of the leaves themselves. Instead of being strongly 

 convex and narrowing rather abruptly from a broad and rounded or 

 subauriculate base, these leaves are plane or only slightly convex 

 and taper more gradually from a narrower and scarcely rounded 

 base. In connection with the var. implexum, Boulay 1 makes the 

 interesting observation that it descends from the higher mountains 

 into the lower woody zone, while the typical form does not descend 

 below the middle woody zone. There is thus a slight difference in 

 altitudinal distribution to support the morphological differences. 



If a species of the character just outlined is recognized, the choice 

 of a name for it is beset with difficulties. " B. triangularis Pears." 

 should apparently be typified by the forma laxa of Schiffner with 

 persistent leaves, but the use of this name in any sense would lead 

 to confusion on account of the older B triangularis Lindb. Nees 

 von Esenbeck's Herpetium deflexum $ implexum was probably a 

 mixture of several forms, but his /S 1 implexum laxius was apparently 

 the same as Miiller's Pleuroschisma, tricrenatum var. implexum. 

 Under /3 1 Nees von Esenbeck, as shown above, included Pleuros- 

 chismafiaccidum Dumort. as a synonym, and Dumortier 's description 2 

 certainly agrees with Miiller's, except that no mention is made of 

 caducous leaves. If it could be established that these were present 



« Muscinees de la France 2: 50. 1904. 

 »Syll. Jung. 71. 1831. 



