664 SUMMABY OF CURRENT RESEARCHES RELATING TO 



branching of the true cyme. This is especially seen from an examination 

 of the double borragoid, like that of Cerinthe. 



Leaf of Taxodium,* — Prof. S. Coulter finds the leaf of Taxodium 

 disticlium to differ from that of Pinus sylvestris (agreeing rather with 

 the young leaves of Pinus) in the following points : — (1) In the less 

 perfect development of the stomates ; (2) in the imperfect development 

 and indefinite arrangement of the strengthening apparatus, shown by the 

 absence of the continuous hypodermal layer and the absence of 

 sclerenchyme from the region of the resin-duct ; (3) in the presence of a 

 single resin-duct, showing imperfect differentiation from the surrounding 

 tissue ; and (4) in the less complete development of the vascular bundle 

 and of its elements. 



Origin of Rootlets.f — In a very elaborate paper on this subject, 

 MM. P. Van Tieghem and H. Douliot state that for the origin of rootlets 

 in Phanerogams, one general and very simple rule may be given. In all 

 Phanerogams the rootlets proceed from a transverse growth localized in 

 the pericycle of the mother-root, and their three regions are cut off in 

 the same manner by two tangential divisions in the group of pericyclic 

 cells which gather together radially. The position of the lateral root 

 varies ; it may be situated either opposite the phloem of a vascular 

 bundle, or opposite a ray, or it may either be placed in the middle of a 

 ray and inserted at the same time on two neighbouring vascular bundles, 

 or laterally and inserted only on the side of the corresponding bundle. 



The existence or absence of a pocket, and its thickness, origin, and 

 mode of separation, are characters which modify the external aspect of 

 the rootlet, but are quite subordinate characters, seeing that they vary 

 not only in allied families, but also in allied genera belonging to the 

 same family, and sometimes from one species to another in the same 

 genus, and occasionally from one root to another in the same plant. 

 Two distinct variations, however, can be traced in the mode of formation 

 of pockets in Phanerogams : either the desquamation of the ej)iderm is 

 partial, as in Dicotyledons (except Nymphfeacese) and Gymnosperms, or 

 the desquamation is total, as in Monocotyledons and Nymphfeacese. 



The authors tben describe the origin and development of lateral 

 roots in Vascular Cryptogams, and state that in this latter group of 

 plants two great divisions can be recognized, the one comprising the 

 FilicineaB, and the other Lycopodium and Isoetes. In the first division 

 the root is formed in the innermost layer of the cortex, that is, in the 

 actual endoderm. Filicinese are, therefore, endodermorhizal and mona- 

 crorhizal. In the second division the root is formed in the outermost 

 layer of the central cylinder, that is, in the pericycle ; Lycopodium and 

 Isoetes are, therefore, pericyclorliizal aud triacrorhi/al. 



If we compare these two groups with Phanerogams, it will be seen 

 that the second resembles them in all points, while the first differs 

 j>reciselj to the same extent as it differs from the second. There are 

 then two types of formation and growth of roots in vascular plants. If 

 we divide Vascular Plants into two great groups, the first would include 

 l^hanerogamp, Lycojpodiuin, aud Isoetes ; while the second would include 

 all Vascular Cryptogams, with the exception of Lycopodium and Isoetes. 



The authors conclude this exceedingly lengthy paper by describing 



* Bot. Gazette, xiv. (1889) pp. 76-81, 101-7 (1 pi. and 1 fig.), 

 t Anil. Sci. Nat., viii. (1888-9) pp. 1-660 (40 pis.). 



