, THE EMBRYO AND THE SEED 277 



defined by the first segmentations of the zygote. Thus the em- 

 bryo of a typical Dicotyledon springs from the two distal cells of 

 the filamentous pro-embryo. The larger part, including the coty- 

 ledons, plumule, hypocotyl, and most of the root arise from the 

 distal cell ; the tip of the root originates from the cell next below 

 it ; the rest of the pro-embryo acts as an organ of attachment. 



While the Capsella-type shows the embryogenv usual in Dicotyledons, 

 aberrant forms are not uncomnaon. But as they are mostly sporadic in their 

 distribution they do not suggest any consistent basis for morphological argu- 

 ment. It is rarely that a family includes many aberrant tj'pes ; an exception 

 is seen in the Leguniinosae, where the peculiarities are most marked in the 

 suspensor. Of the rest, the most interesting variants are the pseudo-mono- 

 cotyledonous embryos. In certain plants that are clearly Dicotyledons in 

 their general characters, only one cotyledon appears. This is seen in Carum 

 bulbocastanum, Eranthis hyemalis and Cyclamen persicuin ; and it is probably 

 due to abortion of one of the cotyledons. But in some cases it is believed to 

 result from a lateral fusion of the two cotyledons to form one, as in Raniin- 

 cidus Ficaria. Much other evidence suggests that the Monocotyledonous 

 state is derivative from that of the Dicot)'ledons. This conclusion is further 

 countenanced by the fact that occasionally Monocotyledons are found with 

 two cotyledons (Agapanthus), or even four growing points may appear on a 

 peripheral zone (Cyrtanthus) . In such cases the apex of the axis would be 

 central and terminal, as in the Capsella-type. But in most Monocotyledons 

 it is lateral, as it is in Alisma. 



The development from the zygote in Alisma or Sagittaria is typical 

 for a large number of Monocotyledons. The first divisions give rise 

 to a three-celled pro-embryo, of which the basal cell [q] is enlarged, 

 and does not divide further (Fig. 219, i.). The middle cell divides 

 into several cells {m, n, 0, p, Fig. 219, ii.). The distal cell divides 

 into quarters, and subsequently by anticlinal and periclinal walls so 

 as to form the single terminal cotyledon (Figs. 11. -v., /). ^leanwhile in 

 the tier of cells below this (Figs. ii. to v., m) a lateral depression begins 

 to appear, which develops into the apex of the axis. This is then 

 lateral m origin, while the cotyledon is terminal. The hypocotyl and 

 root-tip spring from the next two tiers [ii-o, Figs, ii.-v.). It thus appears 

 that the reference of the several parts to cells of the pro-embrj^o 

 dift'ers m the Alisiiia-type from that m the Capsella-type. The 

 most remarkable difference lies in the lateral origin of the apex of the 

 axis in Alisma, while in the Capsella-type the cotyledons are lateral, 

 and the apex distal. 



In a number of Monocotyledons the embrj-o differs from the Alistua-type. 

 The most interesting are those in which the apex of the axis originates from 

 the terminal cell of the pro-embryo. This occurs in the Dioscoreaceae and 



