STRUCTURE OF THE STEM OF THE SCREW PINE. 51 



dicotyledons, as is well known, it is in the fibvo-vascular 

 bundles themselves of the original woody cylinder that the 

 growth takes place. Stems of the Draccena type so far simu- 

 late the woody structure of dicotyledons that the secondary 

 woody growths are arranged more or less in concentric zones, 

 but these growths are altogether independent of the mass of 

 primary bundles. Millardet points out a further distinction 

 in the two cases, in the fact that while the cambial cells in 

 dicotyledons gradually undergo conversion into cells of the 

 particular tissue, whether woody, bast, or vascular, they are 

 destined to add to, in Draccena the new histological elements 

 are fashioned from cells which are produced secondarily by 

 the repeated division of those of particular portions of the 

 cambium layer. 



In monocotyledons generally the fibro-vascular bundles 

 which at their upper extremities pass into the leaves, at their 

 lower terminate obliquely upon the surface of the stem, and 

 are not, therefore, continued through the whole length of the 

 plant's axis. It was originally supposed that the new 

 bundles above mentioned in Draccena were downward pro- 

 longations, subsequently formed, of the primary bundles. This 

 was, in effect, a kind of reminiscence of the old, and now 

 finally dismissed, theory of the descending formation of wood. 

 Millardet, in tracing carefully the development of these 

 secondary bundles, has shown that they have no connection 

 with the original bundles of the stem, and, unlike these, are 

 not inflected inwards towards its axial line, although they 

 are not always free from lateral inclination. 



In consequence of the death of a large and aged screw pine 

 in the Botanic Gardens at Glasnevin from the attacks of a 

 parasitic fungus, which had recently destroyed a similar plant 

 at Breslau, 1 obtained this year a portion of the extremity of 

 one of the branches in a fresh state. After seeing what had 

 been made out by Millardet in the case oi Draccena and Yucca, 

 I was in hopes that I might be able to trace similar facts in 

 Pandanus. As, however, it was wished to attempt to pre- 

 serve the stem of the plant in a dried state as a museum 

 specimen, I was unable to. get any of the older parts to 

 compare with the younger. In the portion I examined I 

 could not satisfy myself thoroughly of the existence of a 

 cambium layer. In the youngest stems, which increase in 

 this way, it does not make its appearance, and this might be 

 the reason of my not finding it in Pandanus. There were, 

 however, several points of considerable interest in the minute 

 structure, and these appear to. me worth description, though 

 only contributions towards a more complete account. 



