172 MORPHOLOGY. 



tissue ; a lower one, which, excepting in RiccieG, is elongated into a 

 pedicle (seta) ; and an upper one, which becomes a spherical sporo- 

 carp (sporocarpium) in Jungermannia pusilla, and a filiform one in 

 Anthoceros. The cellular tissue of this upper part is, again, very 

 variously formed. The most external layers of cells become thick- 

 ened, and form the wall of the sporocarp, which is afterwards torn 

 up from above downward in various ways. In rare cases a portion 

 of the central tissue remains under the form of a central columella, 

 long in Anthoceros, or short as in Pellia epipliylla. Generally speak- 

 ing, it is wholly converted into two different forms of cells : parent 

 cells (in each of which four spores are formed and become clothed 

 by a special membrane), which subsequently become absorbed, and 

 elongated fusiform cells, containing from one to three spiral fibres, 

 and which sometimes appear loose amongst the spores (Fegatella 

 conica), sometimes adhering to the columella (Pellia epiphylla), some- 

 times on the margin (Jungermannia Mcuspidata), or on the point (7. 

 pinguis), or on the inner surface (./. trichophylla) of the valves ; or 

 in rare cases, as in the Ricciece, they are entirely absent. They are 

 termed elaters (elateres). 



The development of the originally homogeneous cellular tissue into 

 such different kinds, that the homogeneous is torn away from the hetero- 

 geneous in consequence of their hygroscopic and elastic properties, occurs 

 here as in the Moss capsule ; and at any rate, during the present imperfect 

 state of our knowledge, we have as little to do in the one case as the 

 other with a separation into original separate parts merely grown together. 

 The manner in which the separation of the parts is effected is very various ; 

 sometimes there only appears a cleft (Monoclea), or the wall is split more 

 or less deeply into valves (valvulce) varying in number from two to eight 

 (Pellia epiphylla^ J. platyphylla, complanata), or into many teeth 

 (denies), more rarely into irregular shreds (Grimaldia hemisphcericd). 

 In more rare cases a separation takes place around the fruit, so that the 

 upper portion falls off as a cover, as in Fimbriaria; in the Ricciece it 

 remains closed until destroyed by external agencies ; in Riccia itself it 

 is absorbed, so that the spores lie free in the cavity of the calyptra. 

 More exact observations probably yet remain to be made concerning the 

 development of the spores. I noticed that there were in the earliest 

 stages always four spores free in the parent cell. I have not hitherto 

 discovered any trace of a division of parent cell by the growing in of a 

 partition wall, as Meyen describes*; but my observations are as yet 

 very incomplete. We have some excellent remarks by Hugo Mohlf on 

 the formation of the spores in Anthoceros Icevis, which appears to cor- 

 respond closely with the formation of the pollen. 



D. The antheridia, whose forms and development entirely cor- 

 respond with those of Mosses, consist of a pedicle, which varies in 

 length, or is entirely wanting ; and of the upper part, which is 

 always spherical or ovate. The leaves rarely form special enve- 

 lopes for these structures, although several leaves are frequently 

 crowded more closely together at the end of the stem, concealing 



* Meyen, Pflanzenphysiologie, iii. p. 391. 

 f Mohi| Linnea, vol. xiii. p. 273. 



