418 Scientific Proceedings, Royal Dublin Society. 



It should be carefully noted that this is a picture of one of the B plants — 

 that is, taken from a leaf which had been grafted on a stem. 



The wood portion (c) gradually extends till a complete ring of secondary 

 wood has been formed in the central bundles. But after the first rapid 

 development the cambial activity becomes quite slow. I have no figure 

 showing the maximum development of the central bundle, even when a plant 

 nearly a yard high was growing on tlie petiole ; but if we imagine at {b) as 

 much secondary wood again as is now present, and at (c) a band about as 

 large as (6) is at present, we have a good idea of its maximum. Why this 

 should be so in every case is hard to explain, because such a development 

 in comparison with the secondary thickenings elsewhere is extremely 

 small. 



Referring to the fig. of the normal petiole, Plate XXX, fig. 1, we 

 remember that on tlie upper side above the central bundle there was an 

 aggregate of bundles. 



Plate XXXII, fig. 2, shows the development of these bundles. Here we 

 see a compact band of secondary tissue, with only one gap at («), and even here 

 we notice that the cells are small and the divisions regular, reminiscent of a 

 cambium formation, {b) and {b) are the primary wood of two adjacent bundles, 

 the rest is secondary development. The figure also shows how with the 

 increased development of wood tlie mechanical ring (c) and (c) becomes 

 ruptured, cortical cells growing into the gaps as at {d). 



Plate XXXII, fig. 3, shows how the compact band is formed. After the 

 fascicular cambium had been active for some time there appeared, as seen in 

 this figure, a series of divisions in the interfascicular cortex. These were 

 not so mathematically regular as the interfascicular divisions in the typical 

 Dicotyledon stem, but they eventually formed a complete line joining up two, 

 three, or in some cases even four bundles of tiie upper surface. This line of 

 cambium gave rise then to a large mass of secondary wood along its whole 

 length, and hence the appearance in Plate XXXII, fig. 2. In this figure 

 we may see a series of divisions at (e) as if the cambium were spreading 

 to other bundles, but it never did so. Never more than four of the upper 

 bundles lying close together were joined up by this interfascicular cambium. 



We have further the peculiar fact that in comparatively few cases was 

 there ever seen a similar joining up of the lower aggregate of bundles, these 

 bundles, as we shall presently see, lagging in their development very much 

 behind the upper ones in tlie majority of cases. 



It was hoped when this interfascicular cambium was first noticed in 

 A, 1353, that all the peripheral bundles would eventually be linked together 

 to form a complete ring resulting in a solid cylinder of wood. But even 



