SECONDARF THICKENING IN MONOCOTYLEDONS. 621 



they are elongated, and spindle-shaped, with their pointed ends pushed between one 

 another in various directions, and provided with oblique slit-like bordered pits (comp. 

 p. i6i) usually arranged in two irregular longitudinal rows on their very thick 

 lignified walls. The peripheral ones adjoin the interfascicular parenchyma without 

 any distinct sheath, and with sparsely pitted surfaces. The number of the tracheides 

 of a transverse section of a bundle is moderate, it may be on the average according 

 to the individual and species as high as 25-60. The form of the whole transverse 

 section of a bundle is more or less broadly elliptical, with the longer axis placed 

 radially ; the relative breadth of the ellipse appears to be inversely proportional to 

 the lateral distance of the bundles one from another, and the average distance seems 

 to differ according to the species. 



In the stems of Aloe and Beaucarnea the structure of the secondary bundles is, 

 according to the data at hand, similar throughout to that described, but it has not 

 yet been investigated with exactitude. The same holds, according to Millardet, for 

 Yucca, with the limitation that the small strand of phloem does not lie in the middle, 

 but at the outer margin of the bundle. 



As has already been stated above, the interfascicular tissue is exclusively 

 parenchyma in the species investigated, that is, not taking into account the crystal- 

 containing sacs, which are scattered in it often in large numbers. Its cells retain on 

 the whole their arrangement in radial rows, in which they passed off from the 

 cambium, though they are often of necessity somewhat displaced around each 

 vascular bundle. Their form also, and especially their height remains, on the whole, 

 similar to that of the cambial cells. In radial direction they ce'-'-ainly undergo 

 a more or less considerable expansion after the division which gave rise to them, 

 so that their transverse section becomes almost quadratic — but rounded off owing to 

 the formation of intercellular spaces ; or it retains the form of a similarly rounded 

 rectangle elongated in a radial direction. This expansion is often specially great in 

 the rows running radially between the sides of closely-grouped bundles, so that in 

 transverse section they resemble for a certain distance here and there the medullary 

 rays with radially-elongated cells found in Dicotyledonous woods; this is very 

 striking, e. g. in Aletris fragrans. 



The structure of the interfascicular xylem-parenchyma shows nothing generally 

 worthy of remark. In the hard woods of the Dracaenas it is provided with rather 

 thick lignified walls, covered with numerous round non-bordered pits, in the other 

 investigated forms it remains thin-walled and sappy. 



It is known that the secondary thickening described continues without limit, 

 and that the old stems of many Dracaenas attain a huge girth as the result of it. 

 It is uncertain how far the necessary periodical remissions and accelerations of 

 growth may lead to inequalities of structure in the wood, corresponding to the 

 formation of annual rings of Dicotyledonous woods. 



The secondary formation of cortex, which is derived from the cambium, is not 

 very extensive, its product is thin-walled cortical parenchyma with crystal-containing 

 sacs. Its cells after leaving the cambial layer undergo single transverse divisions ; 

 in Cordyline paniculata these are frequent, in other forms, such as Calodracon, 

 Aloe sp., and Beaucarnea, they occur here and there ; these cells thus become only 

 half as high as the cambial cells. According to their origin they are at first always 



