350 DR D. H. SCOTT ON THE 



Such isolated tracheides, either between two xylem-strands, or between a xylem- 

 strand and the secondary wood, are not infrequently met with in the transverse 

 sections, as shown in phot. 9, between the strands numbered 14 and 13a in diagram 

 5. It is probable that these elements served to keep up communication between 

 different strands, though in some cases their separation from the bundle to which 

 they belonged may have been an accidental effort of the dilatation of the adjacent 

 parenchyma. 



The wide separation between most of the primary xylem-strands and the secondary 

 wood presents a considerable difficulty, which exists also, though in a less degree, in the 

 case of Calamopitys fascicularis, and probably other species of that genus. In Pitys 

 antiqua, as mentioned above, the actual distance ranges from about a third of a 

 millimetre to almost two millimetres. The average interval, it is true, is only about half 

 a millimetre, but this is equal to twice the average diameter of the xylem-strand itself. 

 Cases of contact are rare, and probably limited to bundles approaching their exit. 

 Apart from these special cases, we find that the number of pith-cells intervening 

 between a primary strand and the wood varies from one up to about twenty, averaging 

 five or six. 



The question arises : What could have been the primary structure of a stem in which 

 such an arrangement prevailed ? If, as all the evidence indicates, the process of 

 secondary growth was of the normal type, the secondary wood being intercalated 

 between the primary xylem and the phloem, it follows that in the primary condition 

 the xylem and phloem of each bundle must have been widely separated, and that to a 

 very unequal extent in different bundles, unless indeed the isolation of the xylem- 

 strands can be explained by subsequent dilatation of the parenchyma. We have 

 already seen that this took place to a considerable extent ; in some cases the apparent 

 doubling of a xylem-strand has evidently been brought about by tangential dilatation of 

 its own fascicular parenchyma, and the isolation of single tracheides may sometimes 

 have been due to a similar cause. But I have found no evidence that in Pitys antiqua 

 the dilatation was greater between the xylem-strand and the secondary wood than else- 

 where. It rather appears that the relative position of the two has remained approxi- 

 mately constant, though the absolute distance between them has no doubt been 

 increased by the general extension of the parenchyma. If this were so, there must 

 have been from the first an unusual separation, varying in different bundles, between 

 xylem and phloem, and the cambium must have arisen towards the phloem-side of the 

 intervening tissue, thus leaving the primary xylem more or less isolated. The tissue 

 between primary and secondary wood would, on this view, have originally been fascicular, 

 but have become assimilated to the pith by subsequent dilatation. The position of the 

 specially deep-seated xylem-strands, behind the outgoing leaf-traces, remains a difficulty, 

 and I do not think that the whole question will be solved until young stems are 

 discovered. Some analogy for the separation of primary from secondary wood is 

 afforded by the peduncles of certain Cycads (especially Stangeria) where the centripetal 



