CHILEAN SPECIES OF METZGERIA. 309 



of time the names "violacea" and " fruticulosa," as applied to Metz- 

 geriae, gradually fell into disuse, and the plants to which they were 

 applied came to be regarded as unimportant forms of M . furcata. 

 Lindberg, fortunately, did not share these views. He considered that 

 R. fruticulosa represented a distinct and well-marked variety of M. 

 furcata and that it was amply distinct from J. violacea, which he 

 regarded as a corresponding variety of M. conjugata. It remained 

 for the writer to restore Dickson's plant to specific rank, under the 

 name M . fruticulosa, and the same recognition, with considerable 

 hesitation, is given to M. violacea in the present paper. 



Through the kindness of Professor Nordstedt of Lund, the writer 

 has had the privilege of studying a part of the type material of Junger- 

 mannia violacea. In spite of its long preservation it soaks up readily 

 in water and retains the vivid coloration to which it owes its name. 

 It lacks sexual organs completely, as the published descriptions 

 emphasize, but shows pointed and highly differentiated gemmiparous 

 branches, to which numerous gemmae still remain attached. These 

 specimens have been carefully compared with the long series of 

 Chilean specimens listed above and do not seem to differ from them 

 in any important respect. M. violacea, as here understood, is exceed- 

 ingly variable; it shows marked reversions, and many of the plants 

 examined exhibit a juvenile or even embryonic stage of development. 

 The following account is therefore somewhat composite in character. 



The more typical vegetative thalli vary from flat to convex, with 

 the margins of the wings more or less revolute. The width is usually 

 0.5-0.8 mm., but narrower thalli are not infrequent and some attain 

 a width of 1 mm. or slightly more. Measured in cells the wings are 

 mostly ten to twenty cells wide. Ventral branching occasionally 

 occurs, although the usual method is by forking, the forks being mostly 

 0.6-1 mm. apart and the entire thallus rarely exceeding a length of 

 0.5-1 cm. The alar cells, taking the mean average from five speci- 

 mens, measure about 31 X 25 n, the highest average obtained being 

 34 X 27 fx and the lowest 27 X 21 /x. The walls are thin throughout 

 and trigones are either absent altogether or minute and inconspicuous. 



The variation in the number and distribution of the hairs is about 

 as great as in M. decipiens and M. epiphylla. In other words a thallus 

 may be hairless throughout the whole or the greater part of its extent, 

 it may produce hairs in abundance, or it may present almost any inter- 

 mediate condition between these extremes. The marginal hairs are 

 not infrequently in pairs but usually arise singly. In fact, on some of 

 the plants examined no twinned hairs could be discovered, although 



