BRITISH HEPATICJ5. 37 



Elaters tubular hyaline, enclosing two brown spiral fibres, xfg-" 

 by ttbV- 



Andrcscium sub-spicate, either seated on distinct shoots (dioi- 

 cous), or on special branches attached to the ventral surface of 

 fertile shoots {autoicous). 



Ferigomal leaves (f. 36, 3) smaller, erecto-patent, more closely 

 imbricated, saccate at the base ; margin entire, nearly plane. 



Antheridia two or three together, axillary, deep green, roundish, 

 shortly stipitate. 



Obs. — N'ordia hyalina aflfords one of those connecting links between genera, so 

 interesting to the philosophic botanist, but, it must be confessed, inconvenient to the 

 systematist. 



Desiring, as far as possible, to simplify the classification, the alternative offered, 

 either to reunite all the species classed under Nardia with Juugermannia (a plan for- 

 merly advocated by Mr. Mitten, but now abandoned) or to choose an intermediate 

 course. 



It seemed impossible to dissever species so closely allied as N. hyalina and 

 N. ohovata, or to class them under any other type than Rardia. 



AUcuki/ria HaskarliaTia, N. ab E. (which is the nearest ally to N. obovata), South- 

 bya tophacea, Spruce, and S. fennica, Gott., should also be placed under this section. 



I cannot refrain from quoting an extract from a letter by Br. Gottsclie (11 Dec. 

 1865), because it shows how careful we should be about judging from indigenous 

 species only : " In my ' Hepaticse Novo-Granatensis ' I have been very puzzled about 

 the difference between Jg. ohovata and Jg. hyalina ; and if I have been able to dis- 

 tinguish the European patches of these plants, as far as I have seen them, I cannot 

 say the same about the American specimens. In Jg. obovata the perianth is really 

 connate with its two folia involucralia, for half their length ; in Jg. hyalina one 

 involucral leaf is commonly free, the other is connate with the lower portion of the 

 perianth. The roots of Jg. obovata are purple and few ; in Jg. hyalina they are more 

 plentiful, and white or red." 



Whether Br. Lindherg's character derived from the inflorescence will prove more 

 decisive time must decide. Certainly N. hyaVma, is not always dioicous. 



Pt. XI. Fig. 36. — N. hyalina. 1. Shoots natural size. 2. Fertile shoot x 8. 

 3. Perigonial leaf and antheridia. 4. Golesule, the lower involucral leaf removed. 

 5. Involucral leafy. 16. 6. Section of colesule below the apex. 7. Section from upper 

 third. 8. Leaf-cells X 120. 



IV. TuicHOCOLEA, Dumort^ 



Jungermannia, Ehrh. in Harm. Mag. 1783, p. 277 ; et Beitr. ii. p. 150 (1788). 



Thricholea, Dumrt. Comm. p. 113(1822). Thricolea, Syll. Jung. p. 66, n. 8 

 (1831) Tricholea, Obs. sur les Jmig. p. 20 (1835). Trichocolea, N. ab E. Leberm. 

 Eur. iii. p. 103 (1838) ; Gottsche in G. & K Hep. Eur. Ex. n. 272 (1863). 



Dioicous. Involucre terminal, or from the growth of innova- 

 tions axillary. Colesule wanting. Oalyptra, when mature, either 

 free, naked, and not incrassate, surrounded only at the base or a 

 little beyond by the abortive pistillidia, and the narrow ring of 



