PRIMARY STRUCTURE OF CERTAIN PALEOZOIC STEMS. 353 



(3) At certain points the primary xylem-strands come into contact with the 

 secondary wood, and pass out through it ; these outgoing strands no doubt represent 

 the leaf-traces. 



(4) The secondary wood is traversed by numerous medullary rays, the larger of 

 which are usually four cells or more thick. The principal rays are much dilated towards 

 their junction with the pith. In addition to the rays, vertical strands of xylem- 

 parenchyma occur, but very sparingly. 



(5) The secondary tracheides have, on their radial walls, several rows of bordered 

 pits, usually contiguous and hexagonal. Tangential pits also occur not infrequently. 

 No true spiral elements are present at the inner edge of the secondary wood. 



The identification of Mr Kidston's specimen, No. 598, on which the description 

 above is based, with the Pitys antiqua of Witham, rests on a comparison with sections 

 of Witham's type-specimen, kindly lent me by Prof. Bayley Balfour, F.R.S., and Mr 

 Kidston. The specimen from which these sections were cut is the large stem shown in 

 transverse section, reduced to half natural size, in Witham's Internal Structure, Plate 

 III. The two sections sent me from Edinburgh are originals of Witham's, while those 

 lent by Mr Kidston (Nos. 217-220 in his Collection) are better and more modern 

 preparations from the same specimen. In all these only the secondary wood is shown. 

 The preservation is not equal to that of the stem (No. 598) collected by Mr Peach, but, 

 allowing for this difference, the structure essentially agrees. The locality, Lennel Braes 

 on the Tweed, is the same. The medullary rays in Witham's specimen are more 

 scattered, and sometimes broader than in No. 598, attaining an extreme width of seven 

 cells, as against five, or rarely six, in the latter, but these differences may well be due to 

 the much greater size of the Witham stem. The dimensions of the elements agree very 

 nearly. Where the pitting is well shown in radial section, the arrangement corresponds 

 with that in No. 598. At many places the pits are in 3-5 rows, covering the whole 

 wall of the tracheide, closely packed and hexagonal in outline, quite like those shown 

 in fig. 18, except that, as the preservation is not so good, the outline of the borders is 

 less sharp. In other places the pits are more scattered and rounded in outline, as also 

 occurs in the other specimen. There is a section in Mr Kidston's collection (No. 221) 

 which, he tells me, may be from a different specimen. This is the section referred to 

 above (p. 346) as showing a pith at least 34 mm. in diameter. The state of preservation 

 is similar to that of the type-specimen, with which its secondary wood exactly agrees. 

 The pith is lacunar, and primary strands of xylem are present, just as in No. 598. A 

 transverse section of a branch, figured by Witham (PI. VII., fig. 11), with a pith more 

 than 3 cm. in diameter, has quite the anatomical habit of our fossil (cf. phot. 5). 



Considering the identical locality, I feel no doubt that the specimen on which the 

 description given in the preceding pages is based, is referable to Witham's species, 

 Pitys antiqua, his 'Lennel Braes Tree.' 



