14 A. G. Tansley. 
The foregoing statements are based on rather a narrow range 
of observations, 1 and in their generalised form certainly admit of 
exceptions in the exact sequence of events; nevertheless they may 
be taken as substantially accurate. It is scarcely necessary to 
point out that the history outlined is exactly that which we have 
seen reason, on comparative grounds, to believe is true of the 
phylogenetic development of the Leptosporangiate Ferns. The 
ontogenetic evidence affords most valuable support to the com¬ 
parative evidence derived from adult structure in reaching our 
phylogenetic conclusions, and the Leptosporangiate Ferns certainly 
furnish a very complete and interesting case of the law of 
recapitulation. 
In the general course of ontogenetic evolution the stele of the 
adult rhizome is arrested at a higher and higher stage of development 
as we pass up the phylogenetic series. It is possible to suppose 
that a kind of reduction may have occurred in certain cases which 
did not involve the actual degeneration of the adult structure, but 
simply the arrest of the ontogenetic evolution at an earlier phase 
than in the immediate ancestors, owing to the requirements of the 
plant having become simpler. Thus we may perhaps explain the 
adult structure in the genus Lindsaya as being a case of arrest of 
the ontogeny at a stage earlier than in possible solenostelic ancestors, 
and this might account for the appearance of such a simple type 
among the Mixta. 
In the case of the Marattiacea we have a very fairly complete 
knowledge of the vascular system in the young plants (Farmer 
and Hill, ’02, Brebner, ’02). Like that of the Leptosporangiate 
Ferns it begins as a protostele, but the subsequent phenomena 
appear, at any rate, to be somewhat different. Instead of the internal 
phloem appearing in the midst of the xylem of the protostele in 
connexion with the internal phloem of the leaf-trace, a stage is 
reached at which the leaf-traces in going off leave the stele cres¬ 
centic in cross section. The crescentic xylem strand sooner or 
later becomes covered on the inside with phloem, pericycle and 
endodermis, and the leaf-traces, at first collateral, later on become 
concentric in structure. The stele sometimes passes through the 
condition of a closed cylinder in the. internodes, but soon becomes 
definitely dictyostelic by the overlapping of leaf-gaps. 
It is not quite easy to know how to interpret this type of 
1 Leclerc du Sablon, ’90, Jeffrey, ’99, ’02, Boodle, ’00, ’01, A. 
Gwynne-Vaughan, ’03, Chandler, ’05, Tansley & Lulham, ’05. 
