Jlepafircr of Norfli Anier'ica. 7 



and this frequently extends to the upper margins, and more 

 rarely to the entire upper surface. Some species of Biccici are 

 whitish, or even milkY white, above. 



True roots are never present, but root-hairs, consisting or- 

 dinarily of a single cell, are usually a])undantly produced on the 

 under surface of the thallus, or, in the foliaceous forms, may 

 proceed from definite points of the leaves (Radiila), or the am- 

 phigastria (FnilJania, Mndotheca), or, as in most, from the 

 under side of the stem, or from both stem and leaves (Junger- 

 mania crenulata). In those forms that live on dry rocks and 

 the bark of trees, the root-hairs are short and fascicled, and 

 are sometimes provided with a sucker-like development at the 

 end. The cell composing the root-hair is usually, in the thal- 

 lose forms, granulose or papillose on the inner surface of its 

 wall. 



Thallus. The thallus is usually dichotomously branched, 

 less frequently somewhat pinnately branched, and in rare cases 

 simple. In some forms it is conspicuously reticulate on the 

 upper surface, and is further marked with large whitish pores 

 ( CouocepliaJiis). 



Leaves. In the foliaceous forms the leaves are usually 

 two-ranked (distichous)^ with frequently a rudimentary row on 

 the ventral surface, known as the amphigastria (Gr. amphi, 

 about, and gasfriou, diminutive of gaster, belly). Both leaves 

 and amphigastria may be entire, serrate, dentate, or variously 

 lobed, cleft or divided. When one of the lobes is much inflated 

 (Fndlania ) it is termed an auricle. The amphigastria usually 

 differ from the leaves more or less in size and shape, though in 

 rare cases they are similar, and the leaves thus become appa- 

 rently three-ranked.* 



Asexual Reproduction. This occurs among the hepat- 

 ics under three forms: viz: (1). By innovations. (2). By 

 gemmae. (3). By runners. 



In nearly all hepatics, except those that are annuals, the 

 growth is continuous and indefinite from the apex of the stems 

 or branches by a process of renewal, while the older portion 



* la it possible that the 3-rankecl condition is the typical form, and 

 that the amphigastria represent the abortive condition resulting from 

 their position on the ventral surface? If so, this would be a marked 

 example of retrograde development. 



