ONTOGENY OF THE CONDUCTING SYSTEM 393 



[Eussow's own names were protoxylem and protophloeni, while 

 De Bary applied the term "Erstlinge" to both classes of ele- 

 ments.] 



In the case of a collateral bundle, the protohadrome and proto- 

 leptome elements are situated at opposite poles in the cross-section of 

 the strand ; the differentiation of the two principal portions of the 

 bundle progresses from these diametrically opposed points in the 

 periphery towards the centre of the strand. In the radial bundles of 

 roots, the first-formed elements of the hadrome- and leptome-plates 

 alternate with one another along the circumference of a circle ; the 

 further differentiation of the strand proceeds centripetally from these 

 starting points. In the concentric bundles of Ferns, finally, the proto- 

 hadrome elements usually occupy the two extremities of the ribbon- 

 shaped area corresponding to the future hadrome-mass. Where the 

 outline of the hadrome is circular or strongly curved, additional points 

 of origin may be found ; as a rule these are uniformly distributed over 

 the cross-section of the strand. 



With regard to the longitudinal differentiation of conducting strands, 

 it may be remarked that acropetal development is characteristic of the 

 bundles of roots, of cauline bundles in stems and also of the leaf- 

 traces of certain plants (Tradescantia albifiora, species of Potamogeton, 

 according to De Bary, and various other Monocotyledons according to 

 Falkenberg). In a very considerable number of Dicotyledons and 

 Conifers, on the other hand, according to Nageli, the leaf-traces develop 

 basipetally. Each vascular strand grows downwards in the stem from 

 a node, and at the same time extends upwards into the leaf to which it 

 belongs. 



Turning, finally, to the development of the individual elements of 

 vascular bundles, we may first of all consider the vessels, which arise by 

 the fusion of longitudinal rows of meristematic cells. Strasburger has 

 shown, in the case of Begonia dioiea and Impatiem glandulosa, that the 

 transverse walls in those rows of cells swell up at an early stage of 

 development, but that they do not become entirely obliterated until the 

 thickening of the longitudinal walls is completed. Indeed a marginal 

 strip of the transverse wall always persists in the form of a narrow 

 annular ridge. The protoplasts of adjacent vessel-segments do not 

 coalesce (Fig. 158). As the thickening of the longitudinal wall pro- 

 gresses, each protoplast dwindles more and more, and finally disappears, 

 nucleus and all. Th. Lange 1S3 asserts that neighbouring protoplasts 

 often fuse after the intervening septum is obliterated (e.g. in Tilin, 

 Mcdva, Hvppwris, Fraxinus, Plantago, GucurMta, Hdionilins); the same 

 observer has shown that in many cases (Cuscuta, Fraxinus, Secalc, 

 Hordeum, Triticum, Pinus laricio, Larix, foliar pulvini of Malva, etc.) 



