1 88 Rhodora [September 



cuticle of L. setacea is distinctly verruculose while that of L. syhatica 

 is smooth or very indistinctly roughened. Unfortunately in slender 

 forms of L. setacea these differences are not always apparent. 



The underleaves of L. sylvatica are usually trifid but are occasion- 

 ally quadrifid on very robust axes and are not infrequently bifid on 

 slender branches. One or two of the divisions are tipped with the 

 remains of hyaline papillae and are thereby aborted in their growth 

 and reduced to one or two cells in length; the divisions without pap- 

 illae become almost as long as the segments of the leaves. In L. 

 setacea quadrifid underleaves are the rule on principal axes, although 

 trifid and even bifid underleaves occur on the branches. Here again 

 the remains of hyaline papillae may be detected on the tips of the 

 divisions; apparently, however, they do not interfere to any great 

 extent with the development of the segments, which never exhibit the 

 extreme disparity in size found in L. sylvatica. Kven on slender 

 forms of Z. setacea this difference in the underleaves seems to be 

 constant. 



The most important differential characters, however, are afforded 

 by the perichaetial leaves. These have been repeatedly figured for 

 L. setacea, but unfortunately the figures show little uniformity. The 

 same statement will also apply to the published descriptions. In 

 Hooker's British Jungermanniae (1816), the perichaetial leaves are 

 figured twice: on plate 8, they appear deeply laciniate with very 

 slender divisions ; on plate i of the supplement, they are ovate in 

 outline and undivided. Nees von Esenbeck ^ comments on plate 8 

 and states that he has never observed the bracts so finely laciniate; 

 Gottsche ^ criticises the same figures and also remarks that those 

 given on plate i are untrue to nature because they represent the 

 bracts as being undivided ; Austin ^ accepts the supplementary fig- 

 ures of the perichaetial leaves but rejects entirely those given on 

 plate 8. As a matter of fact the bracts are almost intermediate in 

 character between the two figures of Hooker ; they are more or less 

 deeply trifid or quadrifid with lanceolate, acuminate, dentate or cili- 

 ate divisions separated by very narrow sinuses. In some cases the 

 primary divisions of the innermost bracts are not very deep, and 

 oftentimes the laciniation seems to be even better marked on unfer- 



' Naturgeschichte der europ. Lel)ermoose, 2: 299. 1836. 

 ^G. & R. Hep. europ. 655. 1879. 

 ^ Hep. Bor.-Amer. yd. 1873. 



