194 SECTIONAL ADDRESSES 



integral portion of the woody axis, but the texture of the wood is still 

 determined by growth characteristics of the cambium which are linked 

 in the closest manner with its origin at the shoot apex, and which are 

 strikingly different in softwoods and hardwoods. 



Conifer and Dicotyledon have very different types of shoot apex. The 

 Conifer bears narrow leaf primordia, many often growing simultaneously 

 at the apex (the seedling often has many cotyledons), and most of the 

 subsequent grovvth of leaf and subtending segment of the axis is in a 

 vertical direction. The Dicotyledon usually has few primordia sharing 

 the growing apex, with a broader leaf primordium, and the seedling has 

 two cotyledons. The subsequent growth of the primordium, whilst 

 mainly longitudinal, also includes considerable tangential expansion. 

 With these differences may be connected the contrast between the long 

 narrow cambium initials of the Conifer (Fig. i), which thoroughly deserve 

 Bailey's term of ' fusiform,' and the shorter Dicotyledon initials which 

 are often not fusiform but more, as Fig. 2 shows, like elongated meristem 

 cells which have retained their original polygonal faces. These charac- 

 teristic forms of the cambial cells have a very distinct bearing upon the 

 differences in the elements cut off from them— differences which affect 

 the grain of the timber and which are well known to all workers in wood, 

 who distinguish sharply between the properties of the softwoods with 

 their uniform grain and freedom from vessels, and the more varied hard- 

 woods with vessels, fibres, etc., variously distributed throughout their 

 texture. 



We will now briefly examine the relation of these structural features to 

 their formation from the cambium, as this tissue resumes activity when 

 growth recommences in the buds. 



Cambium and Vascular Dijferentiation in the Softwood. — In tangential 

 longitudinal view, ends of adjacent fusiform initials never lie at the same 

 level. Evidently as the cambial cylinder grows in size and the initials 

 in the periphery increase in number, new initials do not arise by longi- 

 tudinal radial division. Tangential divisions do not add to the number 

 of cells in the periphery, and the only other divisions that have been seen 

 are transverse divisions, when the two new cells are separated by a some- 

 what oblique cross wall. This wall rapidly assumes a more oblique or 

 vertical position, so that it is usually assumed that the two daughter 

 initials have glided past one another by ' sliding grovrth.' In view of the 

 plastic walls and liquid contents of these meristematic initials it is difficult 

 to understand how they readjust their relative positions by sliding past 

 one another, whilst such a process is also difficult to reconcile with the 

 presence of plasma connections and pits on the radial walls of the vascular 

 elements differentiated from the cambium. Further, if the alteration in 

 the relative position of any two cambial initials with time is followed in 

 the only possible way — viz. by studying the relative displacement of the 

 radial files of tracheids in the woody axis — then such sliding growth does 

 not appear to be a necessary assumption. The tracheids in any radial 

 file, traced outwards through the wood, grow longer but undergo little 

 or no vertical displacement, so that the cambium initial that has been present 

 all along on the outer tangential face of this file has grown longer but 



