DICHOTOMOUS THEORY 137 



based on comparisons as to branching and anatomical structure. These 

 grounds will not suffice to override the inherent probability that the 

 leaves of the Lycopods and Psilotaceae are essentially of the same nature 

 as those of the Sphenophylls or Equiseta, and not the consequence of an 

 entirely distinct evolutionary history. Moreover, on his own admission the 

 " Prohepatic " type, from which Professor Lignier's theory starts, is still wholly 

 hypothetical. Further, it may be remarked that the embryology of the 

 sporophyte gives no assistance to those who would derive it from a 

 dorsiventral thallus. On these as well as other grounds the theory, as 

 stated by Professor Lignier, cannot be upheld. 



An essentially similar hypothesis has been enunciated by Tansley (New 

 Phytologist, 1907, p. 25, etc.). He contemplates a megaphyllous origin 

 of a Fern-like sporophyte from a "hypothetical Archegoniate Alga," which 

 showed dichotomous branching : certain branch-systems became specialised 

 for assimilatory functions as erect shoots, and assumed radial symmetry, 

 while the axis originated by transition through sympodial development of 

 the dichotomy to monopodial branching. On this hypothesis the dorsiventral 

 symmetry would be the primitive and the radial the derivative state in 

 the original sporophyte. The megaphyllous types would be primitive and 

 from these the microphyllous would be derived by widespread reduction. 



Putting aside the collateral speculations of Tansley to which exception 

 may be taken, such as the homoplastic origin of the archegonia and of the 

 spores, as well as of the whole sporophyte in Bryophytes and Pteridophytes, 

 and the wholesale resort to reduction in order to explain the origin 

 of the ancient microphyllous phyla, there are two points of fact, or of 

 absence of fact, which appear specially to oppose his theory : he assumes 

 a radial type of construction to be derivative for the sporophyte [and a 

 dorsiventral type to be primitive ; but in point of fact, in their individual 

 development all sporophytes are originally radial, a condition which has 

 probably a very close relation to their production in the archegonium : 

 that the dorsiventral state is as a general rule derivative in the sporophyte, 

 may be concluded from comparison and shown by experiment (see 

 Chapter XVI.). Further, there is no known case of dichotomy in the 

 sporophyte, where one branch develops as axis and the other as leaf. 

 The known facts derived from living Ferns as well as from the fossils 

 point clearly to dichotomous branching of the axis itself and of the leaf 

 itself, and to transition from a dichotomous to a monopodial branching 

 in the establishment of rachis and pinna. But such evidence is wanting 

 in the relations of leaf and axis. It was chiefly the absence of such 

 evidence that influenced me in rejecting my own suggestion of origin of 

 the shoot from a dichotomous branch-system made in 1884 (Phil. Trans., 

 vol. ii., 1884, p. 605): it applies equally to the theory as stated by 

 Tansley, which appears thus to break down on the test of fact. 



There remains the third view, which, however, is no new one ; for 

 there have not been wanting those who have assigned a more prominent 



