XXX I] 



SORUS 



279 



F'g- 545- -A- leaf of P. pycno- 

 pkylla, itself arrested in its 

 growth, and bearing on its 

 adaxial face a stolon (st). pii 

 = pneumatophores. a, seen in 

 frontal view; b, from the side. 



origin a more advanced state than in Osjnunda. There are thus dififerences 

 of detail, but the underlying similarity is patent, extending as it does to the 

 behaviour on dichotomy (Fig. 544, C). 



In P. pycnopliylla the sections show that occasionally leaves may appear 

 to be substituted by stolons {st, in Fig. 544, A, B), and these may at first 

 have a protostelic vascular supply. But this ex- 

 pands later into a solenostele {st. Fig. 544, A, 

 and Fig. 544, D): later as the stolon bears leaves 

 the structure becomes dictyostelic (see Vol. i, Fig. 

 138). Sometimes a leaf seems to be merged into 

 the stolon, but various relations occur which may 

 be held as depending upon a balance between the 

 leaf-primordium and the stolon-primordium, the 

 resultant structure taking the character of the pre- 

 dominant partner (compare Goebel, Organography, 

 Engl. Edn., Vol. ii, p. 240). This condition re- 

 sembles what has been described by Lang for 

 Osmiinda {Mem. Manch. Lit. and Phil. Vol. 6?>, 

 Part I, p. 53), and such facts have their bearing on the question of the 

 primal relations of axis and leaf (Fig. 545). 



SoRus AND Sporangium 



The drawing by Mettenius (Fig. 542, F) gives a fair idea of the sorus as 

 a whole, with the sporangia borne superficially on the forked veins, a position 

 the same as that seen in Todea barbara. But whereas in the latter the 

 sporangia are formed simultaneously, in Plagiogyria there is a succession 

 of them: not in any gradate sequence, but with different ages intermixed, 

 younger sporangia being interpolated between the older. The interpola- 

 tion is not long continued, and as they mature the sporangia of a sorus 

 appear to be almost of the same age. They are protected by the incurved 

 margins, not by any specialised indusium. The sori are thus superficial, 

 but "mixed." 



The sporangium originates from a single cell, with variable segmentation, 

 like that of Alsophila or Schizaea (Vol. I, Fig. 238, c, d). This provides for 

 a stalk with 5 or 6 rows of cells. The further segmentation is after the 

 Leptosporangiate type, resulting in P. pycnopJiylla in 12 spore-mother-cells, 

 with a typical output of 48 tetrahedral spores. This small output goes, 

 however, with a structure of the sporangial head not highly specialised, while 

 the slit may open either right or left-handed, as in Loxsoniopsis (Fig. 546). 

 The stalk is unusually long for a sporangiinn with oblique annulus, though 

 it is matched by Dicksonia. The head is pear-shaped, as in Loxsoma, and 

 the oblique and rather \ariable ring consists of about 30 cells, surrounding 



