ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 777 



by one onlv, as has been the theory up to the present time. Firstly, when 

 the mother-root possesses more than two woody bundles, and secondly, 

 when the mother-root only possesses two woody bundles. Whenever the 

 structure is binary, the root, whether it be terminal or lateral, or whether 

 it belong to the primary or secondary or any other order, forms its rootlets 

 in the pericycle opposite the intervals which separate the two woody 

 bundles from the toro liber-bundles. 



It is a well-known fact that certain Phanerogams produce regularly 

 buds upon their roots and upon the hypocotylcdonary portion of their 

 stems. These are normal buds, and must not be confused with adventitious 

 buds. Firstly, with regard to the buds which appear shortly after germina- 

 tion on the lower portion of the hypocotylcdonary stem. In order to form 

 one of these buds, three epidermal cells situated at the extremity of the 

 ray which passes between the t^vo liber-bundles and the two woody bundles, 

 divide first by radial, then by tangential and oblique septa, and produce a 

 mass of small cells which forms a projection on the external surface. The 

 bud then is entirely of epidermal origin. The radical buds, on the other 

 hand, are produced at the base of the primary rootlets, and are enclosed 

 within the cortex of the terminal root ; their actual production takes place 

 in the same manner as has been described in the case of hypocotyledonary 

 buds. 



Both radical and hypocotyledonary buds are. therefore, distributed in 

 the root in the same manner as rootlets, and in the stem in the same 

 manner as lateral roots. Frequently they are formed at the same depth as 

 rootlets and lateral roots, that is to say, in the pericycle, and are to the 

 same extent endogenous ; but sometimes, as in Linaria, they are formed in 

 the epidermis, and are exogenous. 



Origin of Rootlets and Lateral Roots in Rubiacese, Violacese, and 

 Apocynacege.* — MM. P. van Tieghem and H. Douliot state that in 

 Eubiaceae the terminal root is binary, and consequently produces its rootlets 

 opposite the intervals which separate the two woody bundles from the 

 two liber-bundles. In Yiolacese the terminal binary root (^Viola nana, 

 V. odarata), or a lateral binary root (V. canadensis^, produces its rootlets 

 in four series in its pericycle. In Apocynaccce the rootlets are formed 

 opposite the woody bundles. In the violet ( V. nana) and in Eubiacete the 

 lateral roots which are produced after germination in the hypocotyledonary 

 stem proceed from the pericycle ; in fact, their origin is the same as that of 

 the rootlets. 



The authors then describe in detail the formation of the lateral roots 

 in Asperula taurina. They conclude the paper by stating that, seeing that 

 the formation of rootlets and early lateral roots in Leguminosas, Cucurb- 

 itacese, Eubiacese, Violacese, and Apocynaceas is found to take place in 

 the usual manner, one can see that only one type of formation for these 

 exists among Dicotyledons. 



In a previous paper f the authors have shown the same to be the case 

 among Monocotyledons ; and it is well known that in Gymnosperms the 

 rootlets are formed in the pericycle of the mother-root. In Vascular 

 Cryptogams it is the endoderm of the root which gives rise to rootlets and 

 lateral roots ; but in this case the endoderm is the external layer of the 

 pericycle. The general conclusion of the authors is that in all vascular 

 plants the rootlets and the early lateral roots are formed in the pericycle of 

 the generating member. 



* Bull. Soc. Bot. France, xxxiv. (1887) pp. 150-4.' 

 + See this Journal, ante, p. 262. 



