284 Scott and Brebner . — On the Anatomy 
nine cells in radial thickness. They are separated from the 
inner margin of the wood by two or three layers of somewhat 
flattened parenchyma resembling that of the pith. The earlier 
stages, however, prove that these cells form part of the original 
procambial strand, and so do not properly belong to the pith 1 . 
Cambial divisions of the cells at the outer edge of these 
phloem-groups go on more actively than in .S. spinosa , and the 
increase of their tissue soon leads to the obliteration of the 
elements lying towards the pith. It is not necessarily the case 
that this obliteration advances regularly from within outwards. 
Often a row of obliterated cells is seen at the extreme inner 
margin ; next we find a row of still normal cells, and then 
another obliterated layer. The older sieve-tubes are evidently 
the first elements to yield to obliteration, while the turgid 
parenchymatous cells offer a longer resistance. As regards 
the primary structure of the medullary phloem, it consists of 
sieve-tubes with somewhat inclined transverse sieve-plates, 
companion cells, and phloem parenchyma, the cells of which 
are about eight diameters in length. The pith itself is of 
uniform structure throughout, consisting of parenchymatous 
cells of variable length, full of starch, with thickened, pitted, 
and somewhat lignified walls. It forms an unyielding matrix 
in which the phloem-groups have no room to expand (see 
Fig. 4)- 
We will complete the history of the medullary phloem before 
going on to the other tissues of the stem. The centrifugal 
activity of the special cambium 2 goes on until a large part of 
each group (sometimes its entire inner half) becomes obliter- 
ated, forming a stratified mass of ‘ Hornbast,’ with no cavities 
left in it beyond mere cracks. The cambial cells can still be 
recognized in Fig. 5 , but their activity has evidently slackened. 
There is no centripetal formation of tissue worth mentioning ; 
1 Cf. Petersen, 1 . c. p. 391. 
2 Vesque calls meristem in this position ‘ false cambium,’ chiefly because its 
productiveness is only unilateral. This is not a constant character, and ‘ special ’ 
or ‘local’ cambium is a less objectionable term. Cf. Vesque, Anatomie com- 
paree de l’ecorce, in Ann. des Sci. Nat. ser. 6, T. ii, 1875, p. 138. 
