400 PRIMARV ARRANGEMENT OF TISSUES. 



in comparative anatomical description, can be brought to bear in its favour. The limi- 

 tation of many vascular bundles by means of the endodermis, to which v/e have adhered 

 in preceding paragraphs, that is to say, the view that the boundary of the bundle is to be 

 drawn at the inner surface of the latter, is, for example, conventional in the sense indi- 

 cated In the bundles of roots this limitation has a genetic basis, in so far as the endo- 

 dermis is the innermost layer of the periblem, while that which it encloses proceeds from 

 the plerome. In other cases, as in the leaves of the Primulas, this developmental ground 

 is absent ; in the case of many bundles of Ferns, developmental considerations in them- 

 selves undoubtedly lead to the opposite result, as the endodermis and the adjoining layer 

 of the bundle are here derived, relatively late, from the division of one layer of 

 mother-cells. Nevertheless, on more extended comparison, the limitation of the bundle 

 on this principle will hardly be disputed. It is further purely conventional, but, as in 

 the example previously adduced, justified by developmental considerations and in the 

 interests of clearness, when the pericambium and the parenchymatous layers within the 

 endodermis of concentric bundles in Ferns are included in the bundle and not treated as 

 separate sheaths. 



The same, or at least quite similar, considerations may be applied to the fibrous 

 strands, the ' bundles of bast-fibres,' which accompany many vascular bundles. There is 

 simply no decisive reason to be found for placing the boundary of a bundle, which they 

 accompany, at their external, or at their internal surface. Where they accompany the 

 vascular bundle, they form with the latter one whole, and the fact that they are com- 

 pleted later than the first, or than many of the elements of the vascular bundle, is in 

 itself no ground for separating them from the bundle, for the essential elements of the 

 latter also attain completion in a definite succession. It is here also conventional, if, as 

 was done above, the fibrous strands are separated from the vascular bundles, and if the 

 boundaries of the latter are drawn on the inner side of the former. The ground for this 

 conventional limitation, however, lies in the fact that the accompanying fibrous strands 

 consist of a sort of tissue different from the essential tissues of all vascular bundles, that 

 they are absent from very many vascular bundles, and are thus non-essential for the 

 vascular bundle generally ; and lastly that they, in a still higher degree than the endo- 

 dermal sheaths, belong to a form of tissue, which in itself forms a system quite inde- 

 pendent of the vascular bundles, and which, in its development, does not stand in any 

 constant relation to the primary differentiation of the meristem. They are portions of a 

 system which may, but which need not, directly accompany the vascular bundles, and they 

 must be included in this system ; it is therefore well to separate them once for all from 

 the vascular bundles. These considerations may also throw light on the discussion 

 which has often been carried on in recent times, and which very clearly illustrates the 

 confusion of ideas now prevailing in the field of anatomy, as to whether the accom- 

 panying fibrous strands belong to the ' fascicular tissue ' or to the 'ground tissue*.' 



The limits between the vascular bundles themselves can be fixed with less difficulty 

 than those between the vascular bundle and its surrounding tissue. If we have once 

 distinguished as a vascular bundle that distinct strand which is formed of tracheae and 

 sieve-tubes definitely grouped, this distinction must be carried out universally, both for 

 the sake of consistency and in the interests of clearness, and every distinct group of the 

 two kinds of organs in question, which forms a united whole, must be called a vascular 

 bundle. What the special grouping of the essential organs may be in these cases, 

 whether the bundle has arisen from the union of several, whether in any one case it 

 corresponds as regards position and origin to a system of numerous bundles which occurs 

 in other cases, these are in themselves questions which have their importance, but they 

 do not touch the anatomical distinction under consideration. On these grounds, the axial 

 bundle of roots, and of the stems of Lycopodiaceae, has been treated as one vascular bundle. 

 It is here also conceded that the subject maybe regarded from other points of view, and thus, 

 for example, we may cease calling the axial bundle of roots a vascular bundle, and regard it 



' [Compare Guillaud, /. c. (see p. 396).] 



