198 SECTIONAL ADDRESSES 



suddenly, in a long series of cells that lie more or less vertically beneath 

 one another, and that it is associated with a very rapid expansion in the 

 size of the future vessel segments. The walls of these elements subse- 

 quently thicken and lignify. Unfortunately the polarising microscope 

 can only tell us about the structure of the thickened wall, and no cross 

 walls are thickened before perforation. In Fraxinus it has been possible 

 to examine cross walls with comparatively small perforations and with 

 a thickened rim. The arrangement of the cellulose micelles in this 

 thickened region certainly suggests that if they were arranged similarly 

 in the original primary wall they would offer minimum resistance to 

 perforation. 



It is interesting to examine more closely the magnitudes involved in 

 the expansion of a hardwood vessel segment as compared with a softwood 

 tracheid. The softwood fusiform initials are so much longer that their 

 volume exceeds that of a hardwood initial , but this volume relation may 

 be reversed during differentiation. Details of the calculations are omitted, 

 but it is estimated that in Scots pine an average cambium cell of 

 length 3-2 mm. has a volume of about 0-00014 cubic mm. In wych elm 

 the cambium cells are about 0-2 mm. long with a volume of about 

 0-000013 cubic mm. The expansion of the tracheid in the pine is almost 

 entirely radial and a spring tracheid may expand to about 7-5 times 

 its original radial diameter, so that its new volume is of the order of 

 0-00105 cubic mm. The vessel segment of the wych elm expands to 

 a roughly circular structure. Quite an average diameter for such a seg- 

 ment would be o - 1 mm., which gives a vessel segment of length o • 2 mm. 

 a volume of 0-0016 cubic mm. Such an average vessel segment may 

 have a volume of the same order of magnitude as a softwood tracheid, 

 but compared with the cell from which it was derived, it has attained 

 its new dimensions by an enormously greater transverse expansion, in 

 this case some 120 times as compared with 7-5 times. 



It seems natural to link this greater expansion with the early collapse of 

 the transverse wall. The collapse of this wall is associated with the fact 

 that a number of cells vertically beneath one another vacuolate almost 

 simultaneously. Possibly the stretching of the wall accelerates the 

 vacuolation of the element immediately beneath in each case, so that the 

 impetus to vacuolate spreads more rapidly downwards beneath a differ- 

 entiating vessel. This would help to explain the other characteristic that 

 distinguishes differentiation in a hardwood from the process in a softwood. 

 In the hardwood the cells formed from the cambium at the same time, and 

 therefore lying in the same tangential plane, do not vacuolate simultane- 

 ously. Some vacuolate before others, and these are always the cells which 

 lie beneath differentiating vessel segments above. But the result is that 

 their expansion can take place at the expense of the plastic elements around 

 them, which are compressed into more elongated elements, and may later 

 differentiate into fibres, or in some cases they vacuolate as they are 

 compressed, as in the oak, where the vessel is surrounded by curiously 

 contorted tracheids. 



In this argument we see the vessel as a natural consequence of 

 the vacuolation of tissue elements which have transverse walls, and 



