FORM OF LEAF-TRACE 193 



or more may pass within a common sheath : a flattened vascular plate, 

 or it may be a curved series of strands, is thus produced. In some leaves 

 of Ferns and in some Dicotyledons the vascular tis.sue thus disposed in a 

 curve as seen in transverse section, with its concave surface upwards, may 

 show a closing in of the lateral margins as the leaf-base is approached; it 

 seems not improbable that this is connected with mechanical requirements 

 consequent on the leverage of. a large leaf on its base. This closing 

 in may even be carried so far that the two edges may become contiguous, 

 and the result will be a structure not unlike that of a cylindrical stele of the 

 axis (Fig. 98). But it would be a mistake, on the mere ground of such 

 structure and without the check of comparison, to suggest any close identity 

 of character of such " pseudosteles " with the stelar condition of the axis. 

 There is reason to believe that these pseudosteles of the leaf are secondary 



FIG. 97. 



Dicksonia Barontetz. Portion of the vascular system of the stem, seen from within, 

 and showing the departure of three leaf-traces. (After Gwynne-Vaughan.) 



in their origin, for it is the fact that they are characteristic of leaves of 

 relatively large size, while smaller leaves are typically dorsiventral in their 

 vascular structure. Moreover, Professor Bertrand and others have been 

 able to show by exact comparative analysis that even in very aberrant 

 cases of Fern-petioles the pseudostelic structure is referable still to a dorsi- 

 ventral origin, and is to be explained as the result of complex foldings 

 and fusions of a band-like vascular tract. A somewhat similar explanation 

 may be given of the "pseudostelic" petioles of such Dicotyledons as are 

 quoted by Schoute (I.e., p 158) and of the "polystelic" petiole of Primula 

 Auricula examined by Gwynne-Vaughan ; these may be held to be secondary 

 modifications of a structure originally dorsiventral, and the position may 

 accordingly be summed up as follows : The construction of the axis is 

 essentially cylindrical, and finds its anatomical expression in the cylindrical 

 stele ; the construction of the leaf is essentially dorsiventral, and it finds 

 its anatomical expression in the isolated vascular strands disposed dorsi- 

 ventrally. Both these are liable to modification in special cases, thus by 



N 



