PRIMARY PARENCnrMA. ROOT. 



4>3 



used is to be called the hypodermal layer, is in most cases equivalent to the outer endo- 

 dermis of the aerial roots of orchids (pp. 125 and 227). Its structure also appears in many 

 roots to be actually that characteristic of the endodermis, as in Elodea according to Nicolai, 

 and also in Acorus Calamus, so that the occurrence of a hypodermal endodermis would 

 thus have a wider distribution than, was stated above. Further investigations of this 

 point are to be undertaken, for the existing data do not appear to me to afford sufficient 

 certainty. The roots of Orchideas investigated also have a denser, smaller-celled paren- 

 chymatous layer below their endodermis, and in the case of most roots the endodermal 

 structure of the hypodermal layer is at least very doubtful. 



Corresponding to the conformity prevailing in the other conditions, and apart 

 from diflferences in the first origination, and in the phenomena of sclerosis to be 

 described later, the structure of the cortex of the root in Filices, Hydropieridea and 

 Equiseta, is similar to that in the Phanerogams. In every sextant of the apical me-, 

 ristem (p. 18) the initial cell of the periblem, which after severance of the plerome- 

 cylinder and the epidermis remains between the two, is divided by a tangential wall 

 into an outer cell which is the initial cell of the outer corlical layer, and an inner initial 

 cell of the inner cortical layer ; the former increases the number of concentric layers 

 by tangential divisions advancing chiefly in centrifugal direction, the latter by those 

 advancing in centripetal direction ; the number of members of each layer increases 

 by radial divisions. In the Filices and Hydropteridese it is the innermost layer of 

 cells, in the Equiseta it is the second layer from inside which becomes converted into 

 the endodermis. In thin roots the tangential divisions are not very numerous, the 

 radial ones in many- cases are even wholly omitted in the inner cortical layer, so that 

 the axial bundle, as seen in cross-section, is only surrounded by six endodermal 

 cells, the latter by the same number of cortical cells, which are relatively very large. 

 (Fig. 169, p. 363.) 



A noticeable phenomenon, which is apparently of frequent occurrence among the 

 Polypodiaceae and Osmundacese, is that the entire cortex soon begins to show dark 

 brown membranes, and uninterrupted connection between them, although the cells 

 do not cease to contain starch, a condition which, however, is by no rrieans common 

 to all Ferns, e. g. it is absent in Scolopendrium, Ophioglossum, Marattiacese, &c. 

 The species of Isoetes and Selaginella and the Lycopodiacese agree in the main with 

 the forms described, as regards the conditions here under consideration. For details 

 and the history of development we may refer to the special works cited. 



The root cap always consists of parenchymatous cells, which in the younger stages 

 contain abundant starch, and are uninterruptedly connected one with another. In old 

 age they die off, as is well known, successively, in many cases after breaking up into 

 layers or rows (exfoliation), a phenomenon which has its anatomical basis in the fact 

 that the limiting layers of the walls become disorganised to an amorphous mucilage. 

 In most roots the number of the successive layers of the cap is large, and the layers 

 as they peel off are constantly replaced by renewed ones, the end of the cap having 

 the conical form represented above, pp. 9 and 12. In other cases the renewal of the 

 layers soon ceases ; e. g. according to Caspary and Nicolai in the Nymphseacese, 

 -^sculus, Najas, Lemna, Hydrilla, and Hyacinthus ; and according to Janczewski in 

 Pistia and Hydrocharis. In the Nymphaeacese, except Nuphar, the cap is persistent, 

 in the other cases mentioned, especially in the thinner, few-layered caps of the 

 water-plants mentioned, the layers successively die and peel off. The main root of 



