904 A TEXTBOOK OF THEORETICAL BOTANY 



growth, and the pattern of their distribution is often an index to the identity 

 of the wood and decides its physical characteristics (Figs. 887 and 888). Two 

 alternative patterns affecting the vessels are so well marked that woods may 

 be classified by them at a glance. In one class the formation of vessels is 

 almost or entirely confined to the spring, so that each annual ring begins with 

 a zone of large vessels. This is called ring-porous wood (Fig. 889). 



In the other class vessels are formed throughout the year. Such wood 

 is called diffuse-porous (Fig. 890). The vessels in these woods are nearly 

 always smaller than in the ring-porous type, and are sometimes scarcely 

 broader than the tracheids. They are also apparently much shorter. The 

 total length of the open lumen in a single vessel is difficult to determine, 

 but the passage of coal-gas through the wood gives us some idea of their 

 relative lengths. Wood without vessels is completely impermeable to gas, 

 but it will pass through a lo-ft. length of the ring-porous wood of Ash in 

 sufficient amount to ignite, while more than 2 ft. of the diifuse-porous wood 

 of the Maple stops it completely. 



A small group of genera belonging to the order Ranales, namely Drimys 

 (Fig. 891), Trochodendron and Tetracentron (Fig. 892), is notable for having 

 no wood vessels at all. These woods are spoken of as homoxylous, for they 

 consist of radially arranged tracheids of uniform width, recalling the wood of 

 Pinus. The resemblance is carried further, for they have large bordered pits 

 on the radial walls of the tracheids, like those of Conifers, though the pitting 

 on the spring wood is scalariform. Whether this is significant from the 

 evolutionary standpoint, as indicating a primitive character in these three 

 genera, or whether they have lost their vessels by reduction in the course 

 of evolution is a matter of opinion. 



Each year's complement of secondary thickening forms one annual 

 ring. The change from the dense, small-celled autumn wood to the soft, 

 open type of spring wood is normally well-marked and unmistakable in all 

 temperate trees, but in trees grown in an equatorial climate growth is con- 

 tinuous and the annual rings may be non-existent. The older fossil woods 

 have either no annual rings or vague rings with no marked break between 

 them, showing that there w^ere no marked seasons even in temperate latitudes 

 in Carboniferous time. A marked differentiation of climate first appears in 

 the Mesozoic period. 



Annual rings are an excellent record of climatic variations, being wider 

 or narrower according to the quality of the season and having a reduced 

 average width in high latitudes and at high altitudes. It does not follow, as 

 might be imagined, that narrow rings imply stronger wood. On the contrary, 

 unfavourable conditions limit chiefly the development of the autumn wood, 

 and narrow rings may consist mainly of the spongy spring wood. The width 

 of the annual ring is not uniform all over the tree but is greater in the upper 

 part, associated with a longer period of cambial activity. 



Annual rings are usually symmetrical in vertical stems, but in oblique 

 or horizontal branches, especially near the base, they become asymmetrical, 

 the lower half (in Conifers) or the upper half (in most Angiosperms) being 



