G48 MR JOHN M'LEAN THOMPSON ON 



sented in figs. 25 and 26. In the first of these the lower surface grooves had died 

 away before the upper grooves, while in the second the upper surface grooves had 

 been lost first. In both cases the infolded margins of the xylem-strap are being 

 reduced, and storage tracheides are appearing in the middle xylem-zone. 



The anatomy of these remarkable filiform leaves has then shown that the trace is 

 of an elaborate type, comparable in form with the trace of the pinnate leaves ; 

 and that, while the trace is admirably constructed for the supplying of extra-marginal 

 pinna-traces, similar to those which feed the ordinary pinnae, no pinna-traces are 

 initiated in the filiform leaves. It may be that the ridges of tissue which run along 

 the margins of the filiform leaves mark the middle zone as of rachial nature. But 

 until definite evidence is advanced, showing that those ridges represent coalesced 

 and aborted pinnae, no definite opinion can be formed regarding the interpretations 

 of the three regions of the filiform leaves. That these leaves are reduction forms 

 seems undeniable ; but whether they are useless abortions or serviceable structures 

 cannot be determined from an herbarium specimen. 



The anatomy of the transitional leaf-forms remains to be considered. Regarding 

 the small filiform leaf with the circinate tip, represented in fig. 3, it will suffice 

 to note that for the greater part of its length its leaf-trace remained of the type 

 shown by the basal region of the longer filiform leaves, and that, as the leaf- tip was 

 approached, the margins of the xylem-strap were reduced and the trace died out. 



The anatomy of the first of the reduction series of four leaves (fig. 5) showed no 

 essential points of difference from an ordinary simply-pinnate leaf. It might be con- 

 sidered a pinnate leaf in miniature. The anatomy of the second leaf of the reduction 

 series (fig. 9) showed two points of interest. In its basal portion it resembled the 

 foregoing leaf, but at a point near to the rachis the petiole became distended. At 

 this point the stele showed a peculiar condition. The xylem-strap was, as usual, 

 divisible into the marginal and middle regions, but a parenchymatous replacement 

 of the tracheides was taking place in the centre of the strap (fig. 27). This led, 

 finally, to what appeared to be an equal division of the stele into two parts. Of these, 

 the one part passed into the pinnate branch, but the other rapidly degenerated, and 

 only an irregular group of tortuous tracheides and a few softer elements entered 

 the aborted branch. I am of the opinion that this leaf might have developed a 

 true dichotomy, but that it became malformed during development. No pinna-trace 

 was found supplying the minute pinna-like protuberances which stood distally on 

 the pinnate branch. 



The terminal portion of the small leaf which provided the third step in the 

 reduction series is worthy of remark (fig. 11). At least five pinnae were supplied with 

 traces. The supplies for the basal pair of pinnae are represented in fig. 34 on their 

 way through the cortex. In fig. 35 the trace to one of the second pair has already 

 been liberated, while the second trace is in process of separation. It is noteworthy 

 that the second pair of traces are here much more bulky than are the first or lower 



