DEVELOPMENT OF THE VASCULAR BUNDLES. 391 



case of numerous Monocotyledons and Dicotyledons, those bundles in any cross- 

 section, which are the first to originate, are also the first to receive the definitive 

 primitive elements of phloem and xylem, does not exclude these possibilities, which 

 are rather to be tested by further investigation. 



So far as is known at present, both origination and completion proceed in the 

 acropetal direction in all root-bundles, in all cauline bundles, and in those in the stem 

 of Filices and Marsiliacese. The leaf-trace bundles of the Phanerogams certainly do 

 not all show the same behaviour. In Tradescantia albiflora, and in species of 

 Potamogeton — with the exception of the lateral bundles of P. crispus — they are both 

 originated and completed in acropetal order of succession, so far as my investigations 

 extend. In Cordyline and Chamsedorea many bundles of the leaf-trace appear, accord- 

 ing to Nageli ', to take the same course of development. Falkenberg ^ makes the 

 same statement for all Monocotyledons investigated by him. 



According to Schmitz ' the median bundle of the leaf-trace of Berberis vulgaris 

 shows the same acropetal progression, at least as regards its origination ; this takes 

 place by means of longitudinal division of a strand of primary meristematic cells, 

 ' starling ' from the point where the median strand of the leaf perpendicularly below 

 (in f^Phyllotaxis, the sixth) curves outwards. The same holds good, according to 

 Nageli *, for the external leaf-trace bundles of Bougainvillea spectabilis, as to which 

 however it is doubtful whether they belong to this series. Frank's statements ^, 

 according to which the same conditions are said to exist in the origination and 

 completion of the trace-bundles in Taxus, Quercus, and JEsculus, require testing. 



On the other hand, it has been shown by Vochting in the case of the bundles of 

 the leaf-trace of the Melastomacese, that with regard to both of the processes in question 

 they grow downwards in the basipetal direction from the point of exit in the node. 

 The same conclusion follows from Sanio's investigations in the year 1864 of the bundles 

 of the PiperacesB, if these, as Weiss " maintains, are bundles of the leaf-trace. The de- 

 finitive completion, at any rate, takes place in very many bundles basipetally from the 

 node of exit; Nageli states this for a portion of the trace-bundles of Chamsedorea and 

 Cordyline, and for those of a very large number of Dicotyledons and Conifers ; and it is 

 not difiicult here, in growing ends of stems, directly to see the basipetally progressive 

 completion of the tracheae; comp. e. g. p. 258, Fig. 115, m^. Thus, as above remarked 

 in demonstrating the course of the bundles in these numerous cases, as we follow them 

 downward from the node, we at the same time follow their order of development. 



The possibility that the development of a leaf-trace bundle may follow different 

 directions in different parts of its course, has been discussed by Mohl,'; in the case of 

 the leaf-trace bundles of the Palm-type, he states that they are develo'ped basipetally 

 I from the node in their upper part, but does not consider that this excludes an 

 acropetal development of the lower part. Neither do Nageli's and Falkenberg's 

 observations, so far as they have been published, appear to me to afford a solution 

 of this question. Development proceeding in two directions is indubitable in the 



' Eeitr. I.e. p. 37. ^ I.e. p. 162. 



' I.e. (see below, p. 393), p. 30. * I.e. p. 121. 



« Botan. Zeitg. 1864, pp. 180, 411. " Cf. p. 250. 



' Verm. Schriften, p. 181. 



