280 EVANS. 



species into the two groups Pinnatae and Furcatae. I nfortunately, as 

 Schiffner has since emphasized (17, p. 1S4), the distinction between 

 these groups is based on variable features, since a thallus may be more 

 or less pinnate in one part and distinctly furcate in another. Even 

 in the more typically pinnate species, such as M. filicina Mitt, of 

 the Andes, M. pubescens and M. frontipilis, the differentiation be- 

 tween an axis and its branches is relatively slight and expresses itself 

 quantitatively rather than qualitatively. In certain species which 

 Stephani includes among the Pinnatae, such as M. decrescens, he 

 utilizes these quantitative differences in determining the relative 

 rank of the branches in a branch -system. In the main axis of M. 

 decrescens, according to his description, the costa is bounded dorsally 

 by five rows of cortical cells and below by seven; in the pinnae (or 

 branches of the first rank) the numbers are four and five, respectively; 

 while in the pinnules (or branches of the second rank) the costa is 

 bounded both dorsally and ventrally by two rows of cortical cells. 

 He describes further a decrease in the width of the wings, correspond- 

 ing with this decrease in the number of cortical cells. Apparently on 

 the basis of these differences he states that the thallus is irregularly 

 pinnately branched and adds that long pinnae and pinnules are mixed 

 with much shorter ones. 



A careful study of the type specimen of M . decrescens in the Mitten 

 Herbarium shows that a pinnate habit is not apparent and that the 

 branches arise in the usual dichotomous manner. These branches, 

 to be sure, show the differences described by Stephani, and there are 

 also intermediate conditions connecting his three types; but these 

 differences do not by any means determine the relative rank of the 

 branches. When an axis forks, for example, either or both of the 

 branches may have a more complex costa and broader wings than the 

 original axis, and the costa of a branch may increase in complexity 

 with its growth in length. The differences are apparently due to- 

 nutritive causes and merely indicate that the costa and wings are 

 variable with respect to the number of their component cells. 



The plants of M. decrescens are whitish or yellowish green and 

 apparently grow in dense mats. The living portion of the thallus may 

 reach a length of 4-5 cm., while the width is mostly 0.9-1.3 mm. The 

 wings are so strongly revolute that their margins sometimes meet, 

 the thallus thus acquiring a cylindrical or subcylindrical form. The 

 successive dichotomies may be as much as 1 cm. apart or as little as 

 1 mm., an individual thallus rarely showing both extremes. Accord- 

 ing to Stephani the ventral surface is naked, the hairs being restricted 



