36 ANATOMY OF THE GYMNOSPERMS 



of the primitive forms has been lost through replacement by or 

 through its being merged by extension and fusion into a growth 

 which is continuous between all points of the cell wall. The 

 capacity for the formation of spirals is thus eliminated from the 

 greater portion of the structure of the wood, though survivals 

 are commonly met with in many of the higher seed plants, where 

 they impart to the wood more or less well defined characters of 

 diagnostic value. This capacity is also lost most completely in 

 the greater number of the Cordaitales, Gingkoales, and Coni- 

 ferales, - - indeed, it may be said that for the secondary growth 

 of the wall it is eventually lost in all species ; but the tendency 

 still survives, a fact made apparent by the observation that in 

 certain species the tertiary layer of the wall invariably gives rise 

 to such spirals, which then constitute definite and reliable diag- 

 nostic features, while in other species they are only imperfectly 

 formed. In all such cases the spirals are to be regarded as sur- 

 vivals, --as the last phases of a tendency which elsewhere has 

 become completely obliterated. 



The origin of the spirals may be traced to a localized second- 

 ary growth of the cell wall. They constitute, in fact, the primi- 

 tive form of the secondary wall which later becomes modified 

 in accordance with altered conditions of growth in such way 

 as to involve an obliteration of the spirals. Such changes are 

 features of progressive development, in consequence of which 

 it is generally true that such structures are always most promi- 

 nent and abundant in the more primitive types, becoming more 

 rare in plants of a higher type of organization and development. 

 Similar relations exist as between the primary and secondary 

 wood of all known Gymnosperms, whence it is possible to recog- 

 nize the general law that spiral tracheids are a feature of the 

 protoxylem, to which region they are wholly confined in the 

 Coniferales and Gingkoales, and almost strictly confined in 

 the Cordaitales, being, with few exceptions among the higher 

 Coniferales, wholly absent from the secondary xylem. 



The direction of the spirals is constant in most cases, being 

 right-handed or ascending from right to left on the side nearest 



