416 XXX. THE SEED OP ANGIOSPERMS. 



covered by the epidermis, shows usually three layers of cortical 

 cells, regularly arranged into a cylindrical sheath, and a median 

 string of elongated cells which runs from the root-apex towards 

 the growing point of the stem. The cortical layers have at the 

 apex only one layer of common initial cells. Over these runs the 

 dermatogen, from which the two layers of root-cap appear seg- 

 mented. The central string, the plerome, is terminated by its own 

 initials. The hypocotyl is prolonged into the single cotyledon, 

 This is bent, corresponding to the form of the seed cavity, tapers 

 gently to its apex, and ends at the chalazal end of the seed. The 

 cotyledon also consists of a cylinder of regularly arranged layers 

 of cells, and is traversed by a central string of elongated cells. 

 This string curves under the plumule, and joins with that of the 

 hypocotyl (see figure). The cell-rows of the cortex also pass over 

 from the hypocotyl into the cotyledon with a gentle curve. The 

 cotyledon in its lower part, like the hypocotyl, has three, higher 

 up, as it tapers, two, and finally only one layer of cortical cells. 

 The central string ends at some little distance from the apex of 

 the cotyledon. In the ripe seed there is not a trace of endosperm. 

 This seed likewise is exalbuminous. The embryo itself has all 

 its cells densely filled with starch. 



The cross- sect'ion of the seed shows nothing new. It presents 

 simultaneously two cross -sections of the embryo, separated by a 

 narrow 7 strip of tissue which passes over into the inner cell-layer 

 of the spermoderm. The structure of the spermoderm is more 

 clear than in longitudinal sections. The cross- sections of the 

 embryo show the concentric arrangement of the cortical layers 

 very beautifully. The division of the pericarp into epi-, meso- 

 and endocarp is even easier to recognise in cross than in longi- 

 tudinal sections ; and the furrow in the middle line of the back is 

 now very noticeable. 



The two angiospermous plants investigated by us offer us 

 truly typical but extreme examples of embryo formation in dicoty- 

 ledons and monocotyledons respectively ; types which are far from 

 exhausting the whole diversity of cases which have been investi- 

 gated. Thus amongst Dicotyledons are examples of embryos 

 which have only one cotyledon (Carum Bulbocastanum, the 

 Pignut, Ranunculus Ficaria, the Pilewort) ; and amongst Mono- 

 cotyledons those in which the cotyledon arises laterally from the 

 growing apex of the stem (Dioscoreaceae, Commelynaceae). Lastly 

 there are seeds in which the embryo persists as a few cells only. 



