THE STRUCTURE AND AFFINITIES OF DIPLOLABIS F.OMERI (SOLMS). 725 



6 '5 ins. and 7 ins. In order to show clearly the early stages of the petiole-trace figured in 

 PI. II. figs. 21 to 24, and PI. III. figs. 30 to 34, I have inserted text-fig. 3, where only the 

 petiole-trace outline, the protoxylem groups, and the islands of parenchyma are indicated. 



We shall now return to PI. IV. fig. 42 and examine the departure of the pinnae from 

 the normal petiole-trace. In that figure two distinct stages of pinna-trace departure 

 are seen. On the left-hand side, the two claw-like projections from the arms are very 

 distinct, whereas, after the departure of a pair of pinna-traces, the extremities of the arms 

 become exactly like those on the left-hand side of PI. II. fig. 29, with blunt ends, though 

 short claws soon appear as on the right-hand side of that figure. 



Since the ends of the arms are enlarged equally in fig. 42, the arms apparently take 

 an equal share in the formation of the pinna-traces. On the right-hand side of fig. 42 

 the claw-like processes have united in the centre. At a, a the line of abscission along 



Fig. 3. —Outline drawings of the early stages of the petiole-trace. The protoxylem groups are shown as black dots. 

 The arrows indicate the path of these protoxylem groups as they come to lie at the ends of the short arms. 



which the bar of xylem b will break away from the arms of the petiole-trace is very 

 distinct. Along the bar itself groups of small cells are easily seen ; these are the proto- 

 xylem strands, and they are arranged in a perfectly symmetrical manner with regard to 

 each other and to the horizontal plane of symmetry of the trace. When the bar be- 

 comes detached from the petiole-trace it breaks up into two and then into four strands. 

 The two inner (PI. IV. fig. 43, pin. tr.), which are much larger than the two outer 

 (aph. tr., aph. tr.), are destined to supply the pinnae. These four strands continue in 

 the petiole cortex for a considerable distance before they diverge into the various organs 

 they supply. In the specimen figured in PI. I. fig. 12, the traces supplying the two 

 pinnae can be recognised quite distinctly at the lower end of the specimen, as is proved 

 by sections cut from that end. This means that they continue to travel upwards in the 

 petiole as distinct, separate strands, without showing any change in form, for nearly 

 three inches before they pass out into the pinnae. In shape the two larger traces 

 resemble that shown in PI. IV. fig. 47, though at first they are rather flatter (PI. IV. 

 fig. 43, pet. tr.). Higher up they become still more curved (PI. IV. fig. 44), and finally, 

 in the pinna, form almost a complete ring. Where the secondary pinna-traces are given 

 off from the primary pinnae, the ring becomes more horseshoe-shaped (PI. IV. fig. 48). 



