— 10— 



archesporium is formed; (v) the presence, in many cases, of stomata on the cap- 

 sule, with accompanying assimilative tissue," — in the following manner, "From 

 the summary presented here, it will be seen that very little importance can be 

 attached to the first and second of the characters mentioned above. The num- 

 ber of chloroplasts per cell varies considerably. The 'endogenous' origin of the 

 antheridia is obviously a secondary charcter, derived from the normal exogenous 

 type of antheridium development; the antheridia themselves agree closely with 

 those of other HepaticcB, especially Sphcerocar pales, and though the development 

 of the antheridial cavity is peculiar, it is easier to relate it to what is found in 

 other HepaticcB than to establish analogies with the Pteridophyta. In the im- 

 mersion of the archegonium in the thallus, the Anthocerotales do not in reality 

 differ widely from other HepaticcE; the archegonium is partly embedded in 

 Aneura, though in this case, the neck is free. The meristematic tissue above the 

 haustorium of the sporogonium in Anthocerotales evidently results from the per- 

 sistence of a stage which is quickly passed through in the development of the 

 sporogonium in other HepaticcB. In Pellia, for instance, the cells of the seta 

 are arranged in regular longitudinal rows, owing to repeated transverse divi- 

 sions in the cells between haustorium and capsule, and we may imagine the sporo- 

 gonium of the Anthocerotales to have arisen through the persistence of the meri- 

 stematic activity of this zone of tissue, the differentiation of the capsule being 

 deferred and becoming basipetal, instead of taking place early and being simul- 

 taneous. The columella of the Anthocerotean capsule, again, ma}^ be compared 

 with the incomplete sterilization of central tissue which results in the formation 

 of the elaterophore of such forms as Pellia, Aneura, and Gottschea splachnophylla. 

 It is doubtful whether much stress should be laid on the presence of stomata in 

 the epidermis of the Anthoceros capsule; this feature is doubtless to be correlated 

 with the development of a many-layered capsule wall consisting of assimilative 

 parenchyma. In various Jungermanniales , the capsule wall is many-layered, 

 and when young contains chloroplasts — a condition which persists in the Antho- 

 cerotales, but is also found in the ripe capsules of forms like Riella, where the cells 

 of the capsule wall do not become thickened with fibres in connection with a 

 special dehiscence mechanism." 



Dr. Cavers does not believe that the Sphagnales any more than the Antho- 

 cerotales should be given co-ordinate rank with Musci and HepaticecB, but sug- 

 gests that if any departure from the customary division into HepaticecB and 

 Musci is to be made, it will be more satisfactory to divide the Bryophytes as a 

 whole into lo groups of co-ordinate rank as given below, 



Sphcerocar pales. 

 Marchantiales. 



Sphagnales. 



Tetraphidales. 



Polytrichales. 



Jungermanniales. 



Anthocerotales. 

 Andreceales. 



Buxbaumiales. 

 Eu-Bryales. 



